Note
Go to the end to download the full example code.
Steel method forward to guess inputs and dynamic shapes (with Tiny-LLM)¶
Inputs are always dynamic with LLMs that is why dynamic shapes
needs to be specified when a LLM is exported with torch.export.export().
Most of the examples on HuggingFace use method
transformers.GenerationMixin.generate() but we only want to
export the model and its method forward.
That example shows to guess the inputs of this method even though the model
is executed through meth generate.
We focus on the model arnir0/Tiny-LLM. To avoid downloading any weights, we write a function creating a random model based on the same architecture.
Steel the forward method¶
The first step is to guess the dummy inputs. Let’s use the true model for that. We use the dummy example from the model page.
import copy
import pprint
import torch
import transformers
from onnx_diagnostic import doc
from onnx_diagnostic.helpers import string_type
from onnx_diagnostic.helpers.torch_helper import steal_forward
from onnx_diagnostic.torch_models.llms import get_tiny_llm
from onnx_diagnostic.torch_export_patches.patch_inputs import use_dyn_not_str
MODEL_NAME = "arnir0/Tiny-LLM"
tokenizer = transformers.AutoTokenizer.from_pretrained(MODEL_NAME)
model = transformers.AutoModelForCausalLM.from_pretrained(MODEL_NAME)
We rewrite the forward method to print the cache dimension.
def _forward_(*args, _f=None, **kwargs):
assert _f is not None
if not hasattr(torch.compiler, "is_exporting") or not torch.compiler.is_exporting():
# torch.compiler.is_exporting requires torch>=2.7
print("<-", string_type((args, kwargs), with_shape=True, with_min_max=True))
res = _f(*args, **kwargs)
if not hasattr(torch.compiler, "is_exporting") or not torch.compiler.is_exporting():
print("->", string_type(res, with_shape=True, with_min_max=True))
return res
keep_model_forward = model.forward
model.forward = lambda *args, _f=keep_model_forward, **kwargs: _forward_(
*args, _f=_f, **kwargs
)
Let’s run the model.
prompt = "Continue: it rains..."
inputs = tokenizer.encode(prompt, return_tensors="pt")
outputs = model.generate(
inputs, max_length=50, temperature=1, top_k=50, top_p=0.95, do_sample=True
)
generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
print("-- prompt", prompt)
print("-- answer", generated_text)
<- ((),dict(cache_position:T7s8[0,7:A3.5],input_ids:T7s1x8[1,29901:A6305.375],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x8x32000[-15.516718864440918,15.75765609741211:A-3.381915190983544],past_key_values:DynamicCache(key_cache=#1[T1s1x1x8x96[-5.490959167480469,6.226877689361572:A-0.11321351693110653]], value_cache=#1[T1s1x1x8x96[-0.6787744760513306,0.49568021297454834:A0.007227749521139988]]))
<- ((),dict(cache_position:T7s1[8,8:A8.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x8x96[-5.490959167480469,6.226877689361572:A-0.11321351693110653]], value_cache=#1[T1s1x1x8x96[-0.6787744760513306,0.49568021297454834:A0.007227749521139988]]),input_ids:T7s1x1[13,13:A13.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-10.432564735412598,8.368535995483398:A-4.234468644971028],past_key_values:DynamicCache(key_cache=#1[T1s1x1x9x96[-5.509540557861328,6.348220348358154:A-0.12195695057461206]], value_cache=#1[T1s1x1x9x96[-0.6787744760513306,0.7704185843467712:A0.009565710057611594]]))
<- ((),dict(cache_position:T7s1[9,9:A9.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x9x96[-5.509540557861328,6.348220348358154:A-0.12195695057461206]], value_cache=#1[T1s1x1x9x96[-0.6787744760513306,0.7704185843467712:A0.009565710057611594]]),input_ids:T7s1x1[29941,29941:A29941.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-18.135845184326172,4.338468074798584:A-9.799261990881526],past_key_values:DynamicCache(key_cache=#1[T1s1x1x10x96[-5.811427116394043,6.348220348358154:A-0.1250392806283344]], value_cache=#1[T1s1x1x10x96[-0.6787744760513306,0.7704185843467712:A0.008353379246409531]]))
<- ((),dict(cache_position:T7s1[10,10:A10.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x10x96[-5.811427116394043,6.348220348358154:A-0.1250392806283344]], value_cache=#1[T1s1x1x10x96[-0.6787744760513306,0.7704185843467712:A0.008353379246409531]]),input_ids:T7s1x1[2610,2610:A2610.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-14.630133628845215,11.018857955932617:A-7.585419899705564],past_key_values:DynamicCache(key_cache=#1[T1s1x1x11x96[-5.811427116394043,6.348220348358154:A-0.12102800812234901]], value_cache=#1[T1s1x1x11x96[-0.6787744760513306,0.7704185843467712:A0.009305229967787565]]))
<- ((),dict(cache_position:T7s1[11,11:A11.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x11x96[-5.811427116394043,6.348220348358154:A-0.12102800812234901]], value_cache=#1[T1s1x1x11x96[-0.6787744760513306,0.7704185843467712:A0.009305229967787565]]),input_ids:T7s1x1[29871,29871:A29871.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-17.729768753051758,15.795294761657715:A-6.686570847850293],past_key_values:DynamicCache(key_cache=#1[T1s1x1x12x96[-5.811427116394043,6.348220348358154:A-0.11321421906611956]], value_cache=#1[T1s1x1x12x96[-0.6787744760513306,0.7704185843467712:A0.0059324780764124325]]))
<- ((),dict(cache_position:T7s1[12,12:A12.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x12x96[-5.811427116394043,6.348220348358154:A-0.11321421906611956]], value_cache=#1[T1s1x1x12x96[-0.6787744760513306,0.7704185843467712:A0.0059324780764124325]]),input_ids:T7s1x1[29906,29906:A29906.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-18.151559829711914,12.831623077392578:A-8.03768296736572],past_key_values:DynamicCache(key_cache=#1[T1s1x1x13x96[-5.811427116394043,6.348220348358154:A-0.09604190427237952]], value_cache=#1[T1s1x1x13x96[-0.6787744760513306,0.7704185843467712:A0.005043171878160371]]))
<- ((),dict(cache_position:T7s1[13,13:A13.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x13x96[-5.811427116394043,6.348220348358154:A-0.09604190427237952]], value_cache=#1[T1s1x1x13x96[-0.6787744760513306,0.7704185843467712:A0.005043171878160371]]),input_ids:T7s1x1[29947,29947:A29947.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-18.921062469482422,10.401754379272461:A-10.987481685367879],past_key_values:DynamicCache(key_cache=#1[T1s1x1x14x96[-5.811427116394043,6.6982550621032715:A-0.08939146100766222]], value_cache=#1[T1s1x1x14x96[-0.6787744760513306,0.7704185843467712:A0.0033156730564877805]]))
<- ((),dict(cache_position:T7s1[14,14:A14.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x14x96[-5.811427116394043,6.6982550621032715:A-0.08939146100766222]], value_cache=#1[T1s1x1x14x96[-0.6787744760513306,0.7704185843467712:A0.0033156730564877805]]),input_ids:T7s1x1[29892,29892:A29892.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-17.109092712402344,11.451603889465332:A-8.615241368453717],past_key_values:DynamicCache(key_cache=#1[T1s1x1x15x96[-5.811427116394043,6.6982550621032715:A-0.08828816101069809]], value_cache=#1[T1s1x1x15x96[-0.6787744760513306,0.7704185843467712:A0.004018123734315143]]))
<- ((),dict(cache_position:T7s1[15,15:A15.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x15x96[-5.811427116394043,6.6982550621032715:A-0.08828816101069809]], value_cache=#1[T1s1x1x15x96[-0.6787744760513306,0.7704185843467712:A0.004018123734315143]]),input_ids:T7s1x1[29871,29871:A29871.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-14.535235404968262,18.55143928527832:A-4.359501328858081],past_key_values:DynamicCache(key_cache=#1[T1s1x1x16x96[-5.811427116394043,6.6982550621032715:A-0.10365996128249814]], value_cache=#1[T1s1x1x16x96[-0.6787744760513306,0.7704185843467712:A0.0018190039553758197]]))
<- ((),dict(cache_position:T7s1[16,16:A16.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x16x96[-5.811427116394043,6.6982550621032715:A-0.10365996128249814]], value_cache=#1[T1s1x1x16x96[-0.6787744760513306,0.7704185843467712:A0.0018190039553758197]]),input_ids:T7s1x1[29906,29906:A29906.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-19.208690643310547,12.71769905090332:A-10.155632720829919],past_key_values:DynamicCache(key_cache=#1[T1s1x1x17x96[-6.856588363647461,6.6982550621032715:A-0.10853694268128795]], value_cache=#1[T1s1x1x17x96[-0.6787744760513306,0.7704185843467712:A0.0013809153403028676]]))
<- ((),dict(cache_position:T7s1[17,17:A17.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x17x96[-6.856588363647461,6.6982550621032715:A-0.10853694268128795]], value_cache=#1[T1s1x1x17x96[-0.6787744760513306,0.7704185843467712:A0.0013809153403028676]]),input_ids:T7s1x1[29900,29900:A29900.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-20.913084030151367,11.392576217651367:A-11.69873876139149],past_key_values:DynamicCache(key_cache=#1[T1s1x1x18x96[-6.856588363647461,6.6982550621032715:A-0.1023646623671084]], value_cache=#1[T1s1x1x18x96[-0.6787744760513306,0.7704185843467712:A-0.00039179591874893014]]))
<- ((),dict(cache_position:T7s1[18,18:A18.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x18x96[-6.856588363647461,6.6982550621032715:A-0.1023646623671084]], value_cache=#1[T1s1x1x18x96[-0.6787744760513306,0.7704185843467712:A-0.00039179591874893014]]),input_ids:T7s1x1[29900,29900:A29900.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-16.21382713317871,13.484505653381348:A-7.445791564540472],past_key_values:DynamicCache(key_cache=#1[T1s1x1x19x96[-6.856588363647461,6.6982550621032715:A-0.08908409716247677]], value_cache=#1[T1s1x1x19x96[-0.6787744760513306,0.7704185843467712:A-0.0019779059926373806]]))
<- ((),dict(cache_position:T7s1[19,19:A19.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x19x96[-6.856588363647461,6.6982550621032715:A-0.08908409716247677]], value_cache=#1[T1s1x1x19x96[-0.6787744760513306,0.7704185843467712:A-0.0019779059926373806]]),input_ids:T7s1x1[29955,29955:A29955.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-21.9020938873291,5.044755458831787:A-13.776848735461943],past_key_values:DynamicCache(key_cache=#1[T1s1x1x20x96[-6.856588363647461,6.840986728668213:A-0.07805764829069327]], value_cache=#1[T1s1x1x20x96[-0.6787744760513306,0.7704185843467712:A-0.00273918923637666]]))
<- ((),dict(cache_position:T7s1[20,20:A20.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x20x96[-6.856588363647461,6.840986728668213:A-0.07805764829069327]], value_cache=#1[T1s1x1x20x96[-0.6787744760513306,0.7704185843467712:A-0.00273918923637666]]),input_ids:T7s1x1[29892,29892:A29892.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-16.70694351196289,7.221324920654297:A-8.892709849011153],past_key_values:DynamicCache(key_cache=#1[T1s1x1x21x96[-6.856588363647461,6.840986728668213:A-0.07859572630424361]], value_cache=#1[T1s1x1x21x96[-0.6787744760513306,0.7704185843467712:A-0.001949111976363571]]))
<- ((),dict(cache_position:T7s1[21,21:A21.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x21x96[-6.856588363647461,6.840986728668213:A-0.07859572630424361]], value_cache=#1[T1s1x1x21x96[-0.6787744760513306,0.7704185843467712:A-0.001949111976363571]]),input_ids:T7s1x1[29871,29871:A29871.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-12.547388076782227,15.494963645935059:A-5.068759219610598],past_key_values:DynamicCache(key_cache=#1[T1s1x1x22x96[-6.856588363647461,6.840986728668213:A-0.07658115250231992]], value_cache=#1[T1s1x1x22x96[-0.6787744760513306,0.7704185843467712:A-0.003277233828743137]]))
<- ((),dict(cache_position:T7s1[22,22:A22.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x22x96[-6.856588363647461,6.840986728668213:A-0.07658115250231992]], value_cache=#1[T1s1x1x22x96[-0.6787744760513306,0.7704185843467712:A-0.003277233828743137]]),input_ids:T7s1x1[29946,29946:A29946.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-22.133716583251953,6.478907585144043:A-11.908872234937736],past_key_values:DynamicCache(key_cache=#1[T1s1x1x23x96[-6.882454872131348,6.840986728668213:A-0.07968059192137224]], value_cache=#1[T1s1x1x23x96[-0.6787744760513306,0.7704185843467712:A-0.003521361922739216]]))
<- ((),dict(cache_position:T7s1[23,23:A23.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x23x96[-6.882454872131348,6.840986728668213:A-0.07968059192137224]], value_cache=#1[T1s1x1x23x96[-0.6787744760513306,0.7704185843467712:A-0.003521361922739216]]),input_ids:T7s1x1[29901,29901:A29901.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-18.94466209411621,7.7586669921875:A-12.086376501435414],past_key_values:DynamicCache(key_cache=#1[T1s1x1x24x96[-6.882454872131348,6.840986728668213:A-0.07993684639399766]], value_cache=#1[T1s1x1x24x96[-0.6787744760513306,0.7704185843467712:A-0.0027230712880800135]]))
<- ((),dict(cache_position:T7s1[24,24:A24.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x24x96[-6.882454872131348,6.840986728668213:A-0.07993684639399766]], value_cache=#1[T1s1x1x24x96[-0.6787744760513306,0.7704185843467712:A-0.0027230712880800135]]),input_ids:T7s1x1[29945,29945:A29945.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-17.87329864501953,9.426261901855469:A-10.317357872662134],past_key_values:DynamicCache(key_cache=#1[T1s1x1x25x96[-6.882454872131348,6.840986728668213:A-0.07471688298367857]], value_cache=#1[T1s1x1x25x96[-0.6787744760513306,0.7704185843467712:A-0.003106062676579313]]))
<- ((),dict(cache_position:T7s1[25,25:A25.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x25x96[-6.882454872131348,6.840986728668213:A-0.07471688298367857]], value_cache=#1[T1s1x1x25x96[-0.6787744760513306,0.7704185843467712:A-0.003106062676579313]]),input_ids:T7s1x1[29953,29953:A29953.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-20.021434783935547,5.666491508483887:A-11.43309489883529],past_key_values:DynamicCache(key_cache=#1[T1s1x1x26x96[-6.882454872131348,6.840986728668213:A-0.06830911466549412]], value_cache=#1[T1s1x1x26x96[-0.6787744760513306,0.7704185843467712:A-0.003428964488013009]]))
<- ((),dict(cache_position:T7s1[26,26:A26.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x26x96[-6.882454872131348,6.840986728668213:A-0.06830911466549412]], value_cache=#1[T1s1x1x26x96[-0.6787744760513306,0.7704185843467712:A-0.003428964488013009]]),input_ids:T7s1x1[13,13:A13.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-8.502959251403809,8.647699356079102:A-3.666581474106759],past_key_values:DynamicCache(key_cache=#1[T1s1x1x27x96[-6.882454872131348,6.840986728668213:A-0.06822447889040902]], value_cache=#1[T1s1x1x27x96[-0.6787744760513306,0.7704185843467712:A-0.00225495119773903]]))
<- ((),dict(cache_position:T7s1[27,27:A27.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x27x96[-6.882454872131348,6.840986728668213:A-0.06822447889040902]], value_cache=#1[T1s1x1x27x96[-0.6787744760513306,0.7704185843467712:A-0.00225495119773903]]),input_ids:T7s1x1[29928,29928:A29928.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-7.725600242614746,10.529221534729004:A-1.5172752494900488],past_key_values:DynamicCache(key_cache=#1[T1s1x1x28x96[-6.882454872131348,6.840986728668213:A-0.06527995995752246]], value_cache=#1[T1s1x1x28x96[-0.6787744760513306,0.7704185843467712:A-0.00138839243755315]]))
<- ((),dict(cache_position:T7s1[28,28:A28.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x28x96[-6.882454872131348,6.840986728668213:A-0.06527995995752246]], value_cache=#1[T1s1x1x28x96[-0.6787744760513306,0.7704185843467712:A-0.00138839243755315]]),input_ids:T7s1x1[485,485:A485.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-7.5223307609558105,12.079378128051758:A-2.200265917599667],past_key_values:DynamicCache(key_cache=#1[T1s1x1x29x96[-6.882454872131348,6.840986728668213:A-0.06148312655263333]], value_cache=#1[T1s1x1x29x96[-0.6787744760513306,0.7704185843467712:A-0.001493190796611151]]))
<- ((),dict(cache_position:T7s1[29,29:A29.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x29x96[-6.882454872131348,6.840986728668213:A-0.06148312655263333]], value_cache=#1[T1s1x1x29x96[-0.6787744760513306,0.7704185843467712:A-0.001493190796611151]]),input_ids:T7s1x1[262,262:A262.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-16.054697036743164,3.195920944213867:A-9.912151685696095],past_key_values:DynamicCache(key_cache=#1[T1s1x1x30x96[-6.882454872131348,6.840986728668213:A-0.0526473199523025]], value_cache=#1[T1s1x1x30x96[-0.6787744760513306,0.7704185843467712:A-0.0016696463759874151]]))
<- ((),dict(cache_position:T7s1[30,30:A30.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x30x96[-6.882454872131348,6.840986728668213:A-0.0526473199523025]], value_cache=#1[T1s1x1x30x96[-0.6787744760513306,0.7704185843467712:A-0.0016696463759874151]]),input_ids:T7s1x1[435,435:A435.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-13.00013542175293,7.802364826202393:A-6.514995807819068],past_key_values:DynamicCache(key_cache=#1[T1s1x1x31x96[-6.882454872131348,6.840986728668213:A-0.04733127825504628]], value_cache=#1[T1s1x1x31x96[-0.6787744760513306,0.7704185843467712:A-0.0008047377704882507]]))
<- ((),dict(cache_position:T7s1[31,31:A31.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x31x96[-6.882454872131348,6.840986728668213:A-0.04733127825504628]], value_cache=#1[T1s1x1x31x96[-0.6787744760513306,0.7704185843467712:A-0.0008047377704882507]]),input_ids:T7s1x1[29892,29892:A29892.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-15.103503227233887,2.7568225860595703:A-9.391665586687624],past_key_values:DynamicCache(key_cache=#1[T1s1x1x32x96[-6.882454872131348,6.840986728668213:A-0.04055644638731337]], value_cache=#1[T1s1x1x32x96[-0.6787744760513306,0.7704185843467712:A-0.0003467011769136737]]))
<- ((),dict(cache_position:T7s1[32,32:A32.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x32x96[-6.882454872131348,6.840986728668213:A-0.04055644638731337]], value_cache=#1[T1s1x1x32x96[-0.6787744760513306,0.7704185843467712:A-0.0003467011769136737]]),input_ids:T7s1x1[322,322:A322.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-16.71291732788086,2.768527030944824:A-9.84994037065329],past_key_values:DynamicCache(key_cache=#1[T1s1x1x33x96[-6.882454872131348,6.840986728668213:A-0.040666309959186665]], value_cache=#1[T1s1x1x33x96[-0.6787744760513306,0.7704185843467712:A-0.0006287981841110609]]))
<- ((),dict(cache_position:T7s1[33,33:A33.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x33x96[-6.882454872131348,6.840986728668213:A-0.040666309959186665]], value_cache=#1[T1s1x1x33x96[-0.6787744760513306,0.7704185843467712:A-0.0006287981841110609]]),input_ids:T7s1x1[390,390:A390.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-12.72878646850586,8.927671432495117:A-5.238189476897475],past_key_values:DynamicCache(key_cache=#1[T1s1x1x34x96[-6.882454872131348,6.840986728668213:A-0.04073903500626758]], value_cache=#1[T1s1x1x34x96[-0.6787744760513306,0.7704185843467712:A-0.00030211803927355217]]))
<- ((),dict(cache_position:T7s1[34,34:A34.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x34x96[-6.882454872131348,6.840986728668213:A-0.04073903500626758]], value_cache=#1[T1s1x1x34x96[-0.6787744760513306,0.7704185843467712:A-0.00030211803927355217]]),input_ids:T7s1x1[29968,29968:A29968.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-17.1890926361084,4.8266496658325195:A-9.480396195146255],past_key_values:DynamicCache(key_cache=#1[T1s1x1x35x96[-6.882454872131348,6.840986728668213:A-0.036860077806375645]], value_cache=#1[T1s1x1x35x96[-0.6787744760513306,0.7704185843467712:A0.00034705521119186806]]))
<- ((),dict(cache_position:T7s1[35,35:A35.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x35x96[-6.882454872131348,6.840986728668213:A-0.036860077806375645]], value_cache=#1[T1s1x1x35x96[-0.6787744760513306,0.7704185843467712:A0.00034705521119186806]]),input_ids:T7s1x1[29928,29928:A29928.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-16.828540802001953,5.075605869293213:A-10.123816122239456],past_key_values:DynamicCache(key_cache=#1[T1s1x1x36x96[-6.882454872131348,6.840986728668213:A-0.03573528471056138]], value_cache=#1[T1s1x1x36x96[-0.6787744760513306,0.7704185843467712:A0.0009487674021994719]]))
<- ((),dict(cache_position:T7s1[36,36:A36.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x36x96[-6.882454872131348,6.840986728668213:A-0.03573528471056138]], value_cache=#1[T1s1x1x36x96[-0.6787744760513306,0.7704185843467712:A0.0009487674021994719]]),input_ids:T7s1x1[29892,29892:A29892.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-15.334418296813965,3.0961825847625732:A-9.918788560789078],past_key_values:DynamicCache(key_cache=#1[T1s1x1x37x96[-6.882454872131348,6.840986728668213:A-0.03620726755868852]], value_cache=#1[T1s1x1x37x96[-0.6787744760513306,0.7704185843467712:A0.0012975151271102485]]))
<- ((),dict(cache_position:T7s1[37,37:A37.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x37x96[-6.882454872131348,6.840986728668213:A-0.03620726755868852]], value_cache=#1[T1s1x1x37x96[-0.6787744760513306,0.7704185843467712:A0.0012975151271102485]]),input_ids:T7s1x1[591,591:A591.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-14.548851013183594,10.435674667358398:A-3.7397975996453314],past_key_values:DynamicCache(key_cache=#1[T1s1x1x38x96[-6.882454872131348,6.840986728668213:A-0.036007717760679826]], value_cache=#1[T1s1x1x38x96[-0.6787744760513306,0.7704185843467712:A0.0013311053766293505]]))
<- ((),dict(cache_position:T7s1[38,38:A38.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x38x96[-6.882454872131348,6.840986728668213:A-0.036007717760679826]], value_cache=#1[T1s1x1x38x96[-0.6787744760513306,0.7704185843467712:A0.0013311053766293505]]),input_ids:T7s1x1[1073,1073:A1073.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-17.840539932250977,6.937159061431885:A-10.19300847663451],past_key_values:DynamicCache(key_cache=#1[T1s1x1x39x96[-6.882454872131348,6.840986728668213:A-0.03415937627835222]], value_cache=#1[T1s1x1x39x96[-0.6787744760513306,0.7704185843467712:A0.00108061570339347]]))
<- ((),dict(cache_position:T7s1[39,39:A39.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x39x96[-6.882454872131348,6.840986728668213:A-0.03415937627835222]], value_cache=#1[T1s1x1x39x96[-0.6787744760513306,0.7704185843467712:A0.00108061570339347]]),input_ids:T7s1x1[29889,29889:A29889.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-14.949212074279785,6.899430274963379:A-8.692059776168783],past_key_values:DynamicCache(key_cache=#1[T1s1x1x40x96[-6.882454872131348,7.078313827514648:A-0.03448053422941181]], value_cache=#1[T1s1x1x40x96[-0.6787744760513306,0.7704185843467712:A0.0013428207877003236]]))
<- ((),dict(cache_position:T7s1[40,40:A40.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x40x96[-6.882454872131348,7.078313827514648:A-0.03448053422941181]], value_cache=#1[T1s1x1x40x96[-0.6787744760513306,0.7704185843467712:A0.0013428207877003236]]),input_ids:T7s1x1[306,306:A306.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-16.425172805786133,9.370534896850586:A-6.1916607732344415],past_key_values:DynamicCache(key_cache=#1[T1s1x1x41x96[-6.882454872131348,7.078313827514648:A-0.03414592427012631]], value_cache=#1[T1s1x1x41x96[-0.6787744760513306,0.7704185843467712:A0.0014931152163145766]]))
<- ((),dict(cache_position:T7s1[41,41:A41.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x41x96[-6.882454872131348,7.078313827514648:A-0.03414592427012631]], value_cache=#1[T1s1x1x41x96[-0.6787744760513306,0.7704185843467712:A0.0014931152163145766]]),input_ids:T7s1x1[29915,29915:A29915.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-12.616035461425781,15.954569816589355:A-4.68313079584483],past_key_values:DynamicCache(key_cache=#1[T1s1x1x42x96[-6.882454872131348,7.078313827514648:A-0.03241555252936941]], value_cache=#1[T1s1x1x42x96[-1.1154754161834717,0.7704185843467712:A0.0008153728593004503]]))
<- ((),dict(cache_position:T7s1[42,42:A42.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x42x96[-6.882454872131348,7.078313827514648:A-0.03241555252936941]], value_cache=#1[T1s1x1x42x96[-1.1154754161834717,0.7704185843467712:A0.0008153728593004503]]),input_ids:T7s1x1[29885,29885:A29885.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-17.34793472290039,5.864483833312988:A-8.361768380252295],past_key_values:DynamicCache(key_cache=#1[T1s1x1x43x96[-6.882454872131348,7.078313827514648:A-0.03249438322728925]], value_cache=#1[T1s1x1x43x96[-1.1154754161834717,0.7704185843467712:A0.0002990090804553964]]))
<- ((),dict(cache_position:T7s1[43,43:A43.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x43x96[-6.882454872131348,7.078313827514648:A-0.03249438322728925]], value_cache=#1[T1s1x1x43x96[-1.1154754161834717,0.7704185843467712:A0.0002990090804553964]]),input_ids:T7s1x1[29871,29871:A29871.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-9.470327377319336,15.861299514770508:A-2.115428224620875],past_key_values:DynamicCache(key_cache=#1[T1s1x1x44x96[-6.882454872131348,7.078313827514648:A-0.028590843711478527]], value_cache=#1[T1s1x1x44x96[-1.1154754161834717,0.7704185843467712:A-0.0004161455061166359]]))
<- ((),dict(cache_position:T7s1[44,44:A44.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x44x96[-6.882454872131348,7.078313827514648:A-0.028590843711478527]], value_cache=#1[T1s1x1x44x96[-1.1154754161834717,0.7704185843467712:A-0.0004161455061166359]]),input_ids:T7s1x1[29947,29947:A29947.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-17.115264892578125,7.146152496337891:A-8.258798646918498],past_key_values:DynamicCache(key_cache=#1[T1s1x1x45x96[-6.882454872131348,7.078313827514648:A-0.028340438807127714]], value_cache=#1[T1s1x1x45x96[-1.1154754161834717,0.7704185843467712:A-0.000832271419875286]]))
<- ((),dict(cache_position:T7s1[45,45:A45.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x45x96[-6.882454872131348,7.078313827514648:A-0.028340438807127714]], value_cache=#1[T1s1x1x45x96[-1.1154754161834717,0.7704185843467712:A-0.000832271419875286]]),input_ids:T7s1x1[29900,29900:A29900.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-18.093294143676758,6.9178547859191895:A-8.125556804327294],past_key_values:DynamicCache(key_cache=#1[T1s1x1x46x96[-6.882454872131348,7.078313827514648:A-0.028565603406525992]], value_cache=#1[T1s1x1x46x96[-1.1154754161834717,0.7704185843467712:A-0.0014778282873264645]]))
<- ((),dict(cache_position:T7s1[46,46:A46.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x46x96[-6.882454872131348,7.078313827514648:A-0.028565603406525992]], value_cache=#1[T1s1x1x46x96[-1.1154754161834717,0.7704185843467712:A-0.0014778282873264645]]),input_ids:T7s1x1[29900,29900:A29900.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-18.948810577392578,3.4617862701416016:A-10.526158623103983],past_key_values:DynamicCache(key_cache=#1[T1s1x1x47x96[-7.198397636413574,7.078313827514648:A-0.028059341673402186]], value_cache=#1[T1s1x1x47x96[-1.1154754161834717,0.7704185843467712:A-0.0020959146497797204]]))
<- ((),dict(cache_position:T7s1[47,47:A47.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x47x96[-7.198397636413574,7.078313827514648:A-0.028059341673402186]], value_cache=#1[T1s1x1x47x96[-1.1154754161834717,0.7704185843467712:A-0.0020959146497797204]]),input_ids:T7s1x1[29900,29900:A29900.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-21.243667602539062,1.9742556810379028:A-12.392495226606727],past_key_values:DynamicCache(key_cache=#1[T1s1x1x48x96[-7.198397636413574,7.078313827514648:A-0.02473806524628546]], value_cache=#1[T1s1x1x48x96[-1.1154754161834717,0.7704185843467712:A-0.002688247413797424]]))
<- ((),dict(cache_position:T7s1[48,48:A48.0],past_key_values:DynamicCache(key_cache=#1[T1s1x1x48x96[-7.198397636413574,7.078313827514648:A-0.02473806524628546]], value_cache=#1[T1s1x1x48x96[-1.1154754161834717,0.7704185843467712:A-0.002688247413797424]]),input_ids:T7s1x1[29941,29941:A29941.0],inputs_embeds:None,use_cache:bool=True,return_dict:bool=True))
-> CausalLMOutputWithPast(logits:T1s1x1x32000[-19.82381820678711,3.515730857849121:A-12.101337580995635],past_key_values:DynamicCache(key_cache=#1[T1s1x1x49x96[-7.198397636413574,7.078313827514648:A-0.01913464772082896]], value_cache=#1[T1s1x1x49x96[-1.1154754161834717,0.7704185843467712:A-0.002685581100340518]]))
-- prompt Continue: it rains...
-- answer Continue: it rains...
3 May 28, 2007, 4:56
Davin J, and RKD, we know. I'm 800036
Let’s restore the forward as it was.
model.forward = keep_model_forward
Another syntax with onnx_diagnostic.helpers.torch_helper.steal_forward().
with steal_forward(model):
model.generate(inputs, max_length=50, temperature=1, top_k=50, top_p=0.95, do_sample=True)
+ -- stolen forward for class LlamaForCausalLM -- iteration 0
<- args=() --- kwargs=dict(cache_position:T7s8,input_ids:T7s1x8,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x8x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x8x96], value_cache=#1[T1s1x1x8x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 1
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x8x96], value_cache=#1[T1s1x1x8x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x9x96], value_cache=#1[T1s1x1x9x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 2
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x9x96], value_cache=#1[T1s1x1x9x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x10x96], value_cache=#1[T1s1x1x10x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 3
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x10x96], value_cache=#1[T1s1x1x10x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x11x96], value_cache=#1[T1s1x1x11x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 4
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x11x96], value_cache=#1[T1s1x1x11x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x12x96], value_cache=#1[T1s1x1x12x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 5
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x12x96], value_cache=#1[T1s1x1x12x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x13x96], value_cache=#1[T1s1x1x13x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 6
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x13x96], value_cache=#1[T1s1x1x13x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x14x96], value_cache=#1[T1s1x1x14x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 7
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x14x96], value_cache=#1[T1s1x1x14x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x15x96], value_cache=#1[T1s1x1x15x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 8
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x15x96], value_cache=#1[T1s1x1x15x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x16x96], value_cache=#1[T1s1x1x16x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 9
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x16x96], value_cache=#1[T1s1x1x16x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x17x96], value_cache=#1[T1s1x1x17x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 10
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x17x96], value_cache=#1[T1s1x1x17x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x18x96], value_cache=#1[T1s1x1x18x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 11
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x18x96], value_cache=#1[T1s1x1x18x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x19x96], value_cache=#1[T1s1x1x19x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 12
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x19x96], value_cache=#1[T1s1x1x19x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x20x96], value_cache=#1[T1s1x1x20x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 13
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x20x96], value_cache=#1[T1s1x1x20x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x21x96], value_cache=#1[T1s1x1x21x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 14
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x21x96], value_cache=#1[T1s1x1x21x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x22x96], value_cache=#1[T1s1x1x22x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 15
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x22x96], value_cache=#1[T1s1x1x22x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x23x96], value_cache=#1[T1s1x1x23x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 16
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x23x96], value_cache=#1[T1s1x1x23x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x24x96], value_cache=#1[T1s1x1x24x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 17
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x24x96], value_cache=#1[T1s1x1x24x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x25x96], value_cache=#1[T1s1x1x25x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 18
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x25x96], value_cache=#1[T1s1x1x25x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x26x96], value_cache=#1[T1s1x1x26x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 19
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x26x96], value_cache=#1[T1s1x1x26x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x27x96], value_cache=#1[T1s1x1x27x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 20
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x27x96], value_cache=#1[T1s1x1x27x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x28x96], value_cache=#1[T1s1x1x28x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 21
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x28x96], value_cache=#1[T1s1x1x28x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x29x96], value_cache=#1[T1s1x1x29x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 22
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x29x96], value_cache=#1[T1s1x1x29x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x30x96], value_cache=#1[T1s1x1x30x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 23
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x30x96], value_cache=#1[T1s1x1x30x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x31x96], value_cache=#1[T1s1x1x31x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 24
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x31x96], value_cache=#1[T1s1x1x31x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x32x96], value_cache=#1[T1s1x1x32x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 25
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x32x96], value_cache=#1[T1s1x1x32x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x33x96], value_cache=#1[T1s1x1x33x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 26
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x33x96], value_cache=#1[T1s1x1x33x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x34x96], value_cache=#1[T1s1x1x34x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 27
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x34x96], value_cache=#1[T1s1x1x34x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x35x96], value_cache=#1[T1s1x1x35x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 28
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x35x96], value_cache=#1[T1s1x1x35x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x36x96], value_cache=#1[T1s1x1x36x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 29
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x36x96], value_cache=#1[T1s1x1x36x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x37x96], value_cache=#1[T1s1x1x37x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 30
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x37x96], value_cache=#1[T1s1x1x37x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x38x96], value_cache=#1[T1s1x1x38x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 31
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x38x96], value_cache=#1[T1s1x1x38x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x39x96], value_cache=#1[T1s1x1x39x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 32
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x39x96], value_cache=#1[T1s1x1x39x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x40x96], value_cache=#1[T1s1x1x40x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 33
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x40x96], value_cache=#1[T1s1x1x40x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x41x96], value_cache=#1[T1s1x1x41x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 34
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x41x96], value_cache=#1[T1s1x1x41x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x42x96], value_cache=#1[T1s1x1x42x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 35
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x42x96], value_cache=#1[T1s1x1x42x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x43x96], value_cache=#1[T1s1x1x43x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 36
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x43x96], value_cache=#1[T1s1x1x43x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x44x96], value_cache=#1[T1s1x1x44x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 37
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x44x96], value_cache=#1[T1s1x1x44x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x45x96], value_cache=#1[T1s1x1x45x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 38
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x45x96], value_cache=#1[T1s1x1x45x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x46x96], value_cache=#1[T1s1x1x46x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 39
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x46x96], value_cache=#1[T1s1x1x46x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x47x96], value_cache=#1[T1s1x1x47x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 40
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x47x96], value_cache=#1[T1s1x1x47x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x48x96], value_cache=#1[T1s1x1x48x96]))
-.
+ -- stolen forward for class LlamaForCausalLM -- iteration 41
<- args=() --- kwargs=dict(cache_position:T7s1,past_key_values:DynamicCache(key_cache=#1[T1s1x1x48x96], value_cache=#1[T1s1x1x48x96]),input_ids:T7s1x1,inputs_embeds:None,use_cache:bool,return_dict:bool)
-> CausalLMOutputWithPast(logits:T1s1x1x32000,past_key_values:DynamicCache(key_cache=#1[T1s1x1x49x96], value_cache=#1[T1s1x1x49x96]))
-.
Untrained model¶
This part can skipped if you are only interested in exporting the original model. It is useful to create a unit test to ensure a specific architecture can be exported despite the many changes brought to torch or transformers.
Let’s create an untrained model using the config file provided
config.json
to create an untrained model:
onnx_diagnostic.torch_models.llms.get_tiny_llm().
Then let’s use it.
experiment = get_tiny_llm()
untrained_model, inputs, dynamic_shapes = (
experiment["model"],
experiment["inputs"],
experiment["dynamic_shapes"],
)
Before we run it, we make a copy of the inputs as the cache get modified by the execution. Then it is no longer valid associated with the previous input_ids and mask.
print("input type before", string_type(inputs, with_shape=True))
expected_output = untrained_model(**inputs)
print("input type after-", string_type(inputs, with_shape=True))
input type before dict(input_ids:T7s2x3,attention_mask:T7s2x33,position_ids:T7s2x3,past_key_values:DynamicCache(key_cache=#1[T1s2x1x30x96], value_cache=#1[T1s2x1x30x96]))
input type after- dict(input_ids:T7s2x3,attention_mask:T7s2x33,position_ids:T7s2x3,past_key_values:DynamicCache(key_cache=#1[T1s2x1x33x96], value_cache=#1[T1s2x1x33x96]))
The outputs
print("result type", string_type(expected_output, with_shape=True))
result type CausalLMOutputWithPast(logits:T1s2x3x32000,past_key_values:DynamicCache(key_cache=#1[T1s2x1x33x96], value_cache=#1[T1s2x1x33x96]))
It works.
ExportedProgram¶
try:
ep = torch.export.export(
untrained_model,
(),
kwargs=cloned_inputs,
dynamic_shapes=use_dyn_not_str(dynamic_shapes),
strict=False,
)
print("It worked:")
print(ep)
except Exception as e:
# To work, it needs at least PRs:
# * https://github.com/huggingface/transformers/pull/36311
# * https://github.com/huggingface/transformers/pull/36652
print("It failed:", e)
It worked:
ExportedProgram:
class GraphModule(torch.nn.Module):
def forward(self, p_model_embed_tokens_weight: "f32[32000, 192]", p_model_layers_0_self_attn_q_proj_weight: "f32[192, 192]", p_model_layers_0_self_attn_k_proj_weight: "f32[96, 192]", p_model_layers_0_self_attn_v_proj_weight: "f32[96, 192]", p_model_layers_0_self_attn_o_proj_weight: "f32[192, 192]", p_model_layers_0_mlp_gate_proj_weight: "f32[1024, 192]", p_model_layers_0_mlp_up_proj_weight: "f32[1024, 192]", p_model_layers_0_mlp_down_proj_weight: "f32[192, 1024]", p_model_layers_0_input_layernorm_weight: "f32[192]", p_model_layers_0_post_attention_layernorm_weight: "f32[192]", p_model_norm_weight: "f32[192]", p_lm_head_weight: "f32[32000, 192]", b_model_rotary_emb_inv_freq: "f32[48]", input_ids: "i64[s44, s70]", attention_mask: "i64[s43, s53]", position_ids: "i64[s44, s70]", past_key_values_key_0: "f32[s44, 1, s45, 96]", past_key_values_value_0: "f32[s44, 1, s21, 96]"):
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:409 in forward, code: causal_mask = create_causal_mask(
function_const_func_spec0 = self.function_const_func_spec0
torch__dynamo__trace_wrapped_higher_order_op_mod_index0 = self.torch__dynamo__trace_wrapped_higher_order_op_ModIndex0
#
sym_size_int_16: "Sym(s70)" = torch.ops.aten.sym_size.int(input_ids, 1)
sym_size_int_19: "Sym(s44)" = torch.ops.aten.sym_size.int(position_ids, 0)
sym_size_int_22: "Sym(s45)" = torch.ops.aten.sym_size.int(past_key_values_key_0, 2)
sym_size_int_24: "Sym(s21)" = torch.ops.aten.sym_size.int(past_key_values_value_0, 2)
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/sparse.py:192 in forward, code: return F.embedding(
embedding: "f32[s44, s70, 192]" = torch.ops.aten.embedding.default(p_model_embed_tokens_weight, input_ids); p_model_embed_tokens_weight = input_ids = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:403 in forward, code: past_seen_tokens, past_seen_tokens + inputs_embeds.shape[1], device=inputs_embeds.device
add: "Sym(s45 + s70)" = sym_size_int_22 + sym_size_int_16
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:402 in forward, code: cache_position: torch.Tensor = torch.arange(
arange: "i64[s70]" = torch.ops.aten.arange.start(sym_size_int_22, add, device = device(type='cpu'), pin_memory = False); sym_size_int_22 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:409 in forward, code: causal_mask = create_causal_mask(
_assert_tensor_metadata_default = torch.ops.aten._assert_tensor_metadata.default(attention_mask, dtype = torch.int64, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default = None
to: "b8[s43, s53]" = torch.ops.aten.to.device(attention_mask, device(type='cpu'), torch.bool); attention_mask = None
arange_1: "i64[s45 + s70]" = torch.ops.aten.arange.default(add, device = device(type='cpu'), pin_memory = False)
add_: "i64[s45 + s70]" = torch.ops.aten.add_.Tensor(arange_1, 0); arange_1 = None
arange_2: "i64[s44]" = torch.ops.aten.arange.default(sym_size_int_19, device = device(type='cpu'), pin_memory = False)
arange_3: "i64[1]" = torch.ops.aten.arange.default(1, device = device(type='cpu'), pin_memory = False)
lazy_load_decompositions = torch._functorch.predispatch.lazy_load_decompositions(); lazy_load_decompositions = None
_vmap_increment_nesting = torch._functorch.predispatch._vmap_increment_nesting(sym_size_int_19, 'error'); _vmap_increment_nesting = None
_add_batch_dim: "i64[]" = torch._functorch.predispatch._add_batch_dim(arange_2, 0, 1); arange_2 = None
lazy_load_decompositions_1 = torch._functorch.predispatch.lazy_load_decompositions(); lazy_load_decompositions_1 = None
_vmap_increment_nesting_1 = torch._functorch.predispatch._vmap_increment_nesting(1, 'error'); _vmap_increment_nesting_1 = None
_add_batch_dim_1: "i64[]" = torch._functorch.predispatch._add_batch_dim(arange_3, 0, 2); arange_3 = _add_batch_dim_1 = None
lazy_load_decompositions_2 = torch._functorch.predispatch.lazy_load_decompositions(); lazy_load_decompositions_2 = None
_vmap_increment_nesting_2 = torch._functorch.predispatch._vmap_increment_nesting(sym_size_int_16, 'error'); _vmap_increment_nesting_2 = None
_add_batch_dim_2: "i64[]" = torch._functorch.predispatch._add_batch_dim(arange, 0, 3); arange = None
lazy_load_decompositions_3 = torch._functorch.predispatch.lazy_load_decompositions(); lazy_load_decompositions_3 = None
_vmap_increment_nesting_3 = torch._functorch.predispatch._vmap_increment_nesting(add, 'error'); _vmap_increment_nesting_3 = None
_add_batch_dim_3: "i64[]" = torch._functorch.predispatch._add_batch_dim(add_, 0, 4); add_ = None
new_ones: "b8[]" = torch.ops.aten.new_ones.default(_add_batch_dim_2, [], dtype = torch.bool, pin_memory = False)
le: "b8[]" = torch.ops.aten.le.Tensor(_add_batch_dim_3, _add_batch_dim_2); _add_batch_dim_2 = None
_assert_tensor_metadata_default_1 = torch.ops.aten._assert_tensor_metadata.default(le, dtype = torch.bool, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_1 = None
to_1: "b8[]" = torch.ops.aten.to.dtype_layout(le, dtype = torch.bool, layout = torch.strided, device = device(type='cpu')); le = None
and_1: "b8[]" = torch.ops.aten.__and__.Tensor(new_ones, to_1); new_ones = to_1 = None
flat_apply: "b8[]" = torch.ops.higher_order.flat_apply(function_const_func_spec0, torch__dynamo__trace_wrapped_higher_order_op_mod_index0, 'torch._dynamo._trace_wrapped_higher_order_op.ModIndex', to, _add_batch_dim, _add_batch_dim_3); function_const_func_spec0 = torch__dynamo__trace_wrapped_higher_order_op_mod_index0 = to = _add_batch_dim = _add_batch_dim_3 = None
_assert_tensor_metadata_default_2 = torch.ops.aten._assert_tensor_metadata.default(flat_apply, dtype = torch.bool, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_2 = None
to_2: "b8[]" = torch.ops.aten.to.dtype_layout(flat_apply, dtype = torch.bool, layout = torch.strided, device = device(type='cpu')); flat_apply = None
and_2: "b8[]" = torch.ops.aten.__and__.Tensor(and_1, to_2); and_1 = to_2 = None
_remove_batch_dim: "b8[s45 + s70]" = torch._functorch.predispatch._remove_batch_dim(and_2, 4, add, 0); and_2 = None
_vmap_decrement_nesting = torch._functorch.predispatch._vmap_decrement_nesting(); _vmap_decrement_nesting = None
_remove_batch_dim_1: "b8[s70, s45 + s70]" = torch._functorch.predispatch._remove_batch_dim(_remove_batch_dim, 3, sym_size_int_16, 0); _remove_batch_dim = None
_vmap_decrement_nesting_1 = torch._functorch.predispatch._vmap_decrement_nesting(); _vmap_decrement_nesting_1 = None
_remove_batch_dim_2: "b8[1, s70, s45 + s70]" = torch._functorch.predispatch._remove_batch_dim(_remove_batch_dim_1, 2, 1, 0)
expand: "b8[1, s70, s45 + s70]" = torch.ops.aten.expand.default(_remove_batch_dim_1, [1, sym_size_int_16, add]); _remove_batch_dim_1 = expand = None
_vmap_decrement_nesting_2 = torch._functorch.predispatch._vmap_decrement_nesting(); _vmap_decrement_nesting_2 = None
_remove_batch_dim_3: "b8[s44, 1, s70, s45 + s70]" = torch._functorch.predispatch._remove_batch_dim(_remove_batch_dim_2, 1, sym_size_int_19, 0); _remove_batch_dim_2 = None
_vmap_decrement_nesting_3 = torch._functorch.predispatch._vmap_decrement_nesting(); _vmap_decrement_nesting_3 = None
# No stacktrace found for following nodes
submod_3 = self.submod_1
wrap_with_set_grad_enabled = torch.ops.higher_order.wrap_with_set_grad_enabled(False, submod_3, b_model_rotary_emb_inv_freq, sym_size_int_19, position_ids); submod_3 = b_model_rotary_emb_inv_freq = position_ids = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:135 in forward, code: return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype)
to_8: "f32[s44, s70, 96]" = wrap_with_set_grad_enabled[0]
to_9: "f32[s44, s70, 96]" = wrap_with_set_grad_enabled[1]; wrap_with_set_grad_enabled = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:64 in forward, code: hidden_states = hidden_states.to(torch.float32)
_assert_tensor_metadata_default_10 = torch.ops.aten._assert_tensor_metadata.default(embedding, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_10 = None
to_10: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(embedding, torch.float32); embedding = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:65 in forward, code: variance = hidden_states.pow(2).mean(-1, keepdim=True)
pow_1: "f32[s44, s70, 192]" = torch.ops.aten.pow.Tensor_Scalar(to_10, 2)
mean: "f32[s44, s70, 1]" = torch.ops.aten.mean.dim(pow_1, [-1], True); pow_1 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:66 in forward, code: hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
add_4: "f32[s44, s70, 1]" = torch.ops.aten.add.Tensor(mean, 1e-05); mean = None
rsqrt: "f32[s44, s70, 1]" = torch.ops.aten.rsqrt.default(add_4); add_4 = None
mul_7: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(to_10, rsqrt); rsqrt = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:67 in forward, code: return self.weight * hidden_states.to(input_dtype)
_assert_tensor_metadata_default_11 = torch.ops.aten._assert_tensor_metadata.default(mul_7, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_11 = None
to_11: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(mul_7, torch.float32); mul_7 = None
mul_8: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(p_model_layers_0_input_layernorm_weight, to_11); p_model_layers_0_input_layernorm_weight = to_11 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear: "f32[s44, s70, 192]" = torch.ops.aten.linear.default(mul_8, p_model_layers_0_self_attn_q_proj_weight); p_model_layers_0_self_attn_q_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:264 in forward, code: query_states = self.q_proj(hidden_states).view(hidden_shape).transpose(1, 2)
view: "f32[s44, s70, 2, 96]" = torch.ops.aten.view.default(linear, [sym_size_int_19, sym_size_int_16, -1, 96]); linear = None
transpose_1: "f32[s44, 2, s70, 96]" = torch.ops.aten.transpose.int(view, 1, 2); view = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_1: "f32[s44, s70, 96]" = torch.ops.aten.linear.default(mul_8, p_model_layers_0_self_attn_k_proj_weight); p_model_layers_0_self_attn_k_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:265 in forward, code: key_states = self.k_proj(hidden_states).view(hidden_shape).transpose(1, 2)
view_1: "f32[s44, s70, 1, 96]" = torch.ops.aten.view.default(linear_1, [sym_size_int_19, sym_size_int_16, -1, 96]); linear_1 = None
transpose_2: "f32[s44, 1, s70, 96]" = torch.ops.aten.transpose.int(view_1, 1, 2); view_1 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_2: "f32[s44, s70, 96]" = torch.ops.aten.linear.default(mul_8, p_model_layers_0_self_attn_v_proj_weight); mul_8 = p_model_layers_0_self_attn_v_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:266 in forward, code: value_states = self.v_proj(hidden_states).view(hidden_shape).transpose(1, 2)
view_2: "f32[s44, s70, 1, 96]" = torch.ops.aten.view.default(linear_2, [sym_size_int_19, sym_size_int_16, -1, 96]); linear_2 = None
transpose_3: "f32[s44, 1, s70, 96]" = torch.ops.aten.transpose.int(view_2, 1, 2); view_2 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:269 in forward, code: query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin)
unsqueeze_3: "f32[s44, 1, s70, 96]" = torch.ops.aten.unsqueeze.default(to_8, 1); to_8 = None
unsqueeze_4: "f32[s44, 1, s70, 96]" = torch.ops.aten.unsqueeze.default(to_9, 1); to_9 = None
mul_9: "f32[s44, 2, s70, 96]" = torch.ops.aten.mul.Tensor(transpose_1, unsqueeze_3)
slice_3: "f32[s44, 2, s70, 48]" = torch.ops.aten.slice.Tensor(transpose_1, 3, 0, 48)
slice_4: "f32[s44, 2, s70, 48]" = torch.ops.aten.slice.Tensor(transpose_1, 3, 48, 9223372036854775807); transpose_1 = None
neg: "f32[s44, 2, s70, 48]" = torch.ops.aten.neg.default(slice_4); slice_4 = None
cat_1: "f32[s44, 2, s70, 96]" = torch.ops.aten.cat.default([neg, slice_3], -1); neg = slice_3 = None
mul_10: "f32[s44, 2, s70, 96]" = torch.ops.aten.mul.Tensor(cat_1, unsqueeze_4); cat_1 = None
add_5: "f32[s44, 2, s70, 96]" = torch.ops.aten.add.Tensor(mul_9, mul_10); mul_9 = mul_10 = None
mul_11: "f32[s44, 1, s70, 96]" = torch.ops.aten.mul.Tensor(transpose_2, unsqueeze_3); unsqueeze_3 = None
slice_5: "f32[s44, 1, s70, 48]" = torch.ops.aten.slice.Tensor(transpose_2, 3, 0, 48)
slice_6: "f32[s44, 1, s70, 48]" = torch.ops.aten.slice.Tensor(transpose_2, 3, 48, 9223372036854775807); transpose_2 = None
neg_1: "f32[s44, 1, s70, 48]" = torch.ops.aten.neg.default(slice_6); slice_6 = None
cat_2: "f32[s44, 1, s70, 96]" = torch.ops.aten.cat.default([neg_1, slice_5], -1); neg_1 = slice_5 = None
mul_12: "f32[s44, 1, s70, 96]" = torch.ops.aten.mul.Tensor(cat_2, unsqueeze_4); cat_2 = unsqueeze_4 = None
add_6: "f32[s44, 1, s70, 96]" = torch.ops.aten.add.Tensor(mul_11, mul_12); mul_11 = mul_12 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:274 in forward, code: key_states, value_states = past_key_values.update(key_states, value_states, self.layer_idx, cache_kwargs)
cat_3: "f32[s44, 1, s45 + s70, 96]" = torch.ops.aten.cat.default([past_key_values_key_0, add_6], -2); past_key_values_key_0 = add_6 = None
cat_4: "f32[s44, 1, s21 + s70, 96]" = torch.ops.aten.cat.default([past_key_values_value_0, transpose_3], -2); past_key_values_value_0 = transpose_3 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:280 in forward, code: attn_output, attn_weights = attention_interface(
slice_7: "f32[s44, 1, s45 + s70, 96]" = torch.ops.aten.slice.Tensor(cat_3, 0, 0, 9223372036854775807)
unsqueeze_5: "f32[s44, 1, 1, s45 + s70, 96]" = torch.ops.aten.unsqueeze.default(slice_7, 2); slice_7 = None
slice_8: "f32[s44, 1, 1, s45 + s70, 96]" = torch.ops.aten.slice.Tensor(unsqueeze_5, 3, 0, 9223372036854775807); unsqueeze_5 = None
expand_2: "f32[s44, 1, 2, s45 + s70, 96]" = torch.ops.aten.expand.default(slice_8, [sym_size_int_19, 1, 2, add, 96]); slice_8 = None
reshape: "f32[s44, 2, s45 + s70, 96]" = torch.ops.aten.reshape.default(expand_2, [sym_size_int_19, 2, add, 96]); expand_2 = None
slice_9: "f32[s44, 1, s21 + s70, 96]" = torch.ops.aten.slice.Tensor(cat_4, 0, 0, 9223372036854775807)
unsqueeze_6: "f32[s44, 1, 1, s21 + s70, 96]" = torch.ops.aten.unsqueeze.default(slice_9, 2); slice_9 = None
add_11: "Sym(s21 + s70)" = sym_size_int_24 + sym_size_int_16; sym_size_int_24 = None
slice_10: "f32[s44, 1, 1, s21 + s70, 96]" = torch.ops.aten.slice.Tensor(unsqueeze_6, 3, 0, 9223372036854775807); unsqueeze_6 = None
expand_3: "f32[s44, 1, 2, s21 + s70, 96]" = torch.ops.aten.expand.default(slice_10, [sym_size_int_19, 1, 2, add_11, 96]); slice_10 = None
reshape_1: "f32[s44, 2, s21 + s70, 96]" = torch.ops.aten.reshape.default(expand_3, [sym_size_int_19, 2, add_11, 96]); expand_3 = add_11 = None
slice_11: "b8[s44, 1, s70, s45 + s70]" = torch.ops.aten.slice.Tensor(_remove_batch_dim_3, 3, None, add); _remove_batch_dim_3 = add = None
scaled_dot_product_attention: "f32[s44, 2, s70, 96]" = torch.ops.aten.scaled_dot_product_attention.default(add_5, reshape, reshape_1, slice_11, scale = 0.10206207261596575); add_5 = reshape = reshape_1 = slice_11 = None
transpose_4: "f32[s44, s70, 2, 96]" = torch.ops.aten.transpose.int(scaled_dot_product_attention, 1, 2); scaled_dot_product_attention = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:291 in forward, code: attn_output = attn_output.reshape(*input_shape, -1).contiguous()
reshape_2: "f32[s44, s70, 192]" = torch.ops.aten.reshape.default(transpose_4, [sym_size_int_19, sym_size_int_16, -1]); transpose_4 = sym_size_int_19 = sym_size_int_16 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_3: "f32[s44, s70, 192]" = torch.ops.aten.linear.default(reshape_2, p_model_layers_0_self_attn_o_proj_weight); reshape_2 = p_model_layers_0_self_attn_o_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:331 in forward, code: hidden_states = residual + hidden_states
add_7: "f32[s44, s70, 192]" = torch.ops.aten.add.Tensor(to_10, linear_3); to_10 = linear_3 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:64 in forward, code: hidden_states = hidden_states.to(torch.float32)
_assert_tensor_metadata_default_12 = torch.ops.aten._assert_tensor_metadata.default(add_7, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_12 = None
to_12: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(add_7, torch.float32); add_7 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:65 in forward, code: variance = hidden_states.pow(2).mean(-1, keepdim=True)
pow_2: "f32[s44, s70, 192]" = torch.ops.aten.pow.Tensor_Scalar(to_12, 2)
mean_1: "f32[s44, s70, 1]" = torch.ops.aten.mean.dim(pow_2, [-1], True); pow_2 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:66 in forward, code: hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
add_8: "f32[s44, s70, 1]" = torch.ops.aten.add.Tensor(mean_1, 1e-05); mean_1 = None
rsqrt_1: "f32[s44, s70, 1]" = torch.ops.aten.rsqrt.default(add_8); add_8 = None
mul_21: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(to_12, rsqrt_1); rsqrt_1 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:67 in forward, code: return self.weight * hidden_states.to(input_dtype)
_assert_tensor_metadata_default_13 = torch.ops.aten._assert_tensor_metadata.default(mul_21, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_13 = None
to_13: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(mul_21, torch.float32); mul_21 = None
mul_22: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(p_model_layers_0_post_attention_layernorm_weight, to_13); p_model_layers_0_post_attention_layernorm_weight = to_13 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_4: "f32[s44, s70, 1024]" = torch.ops.aten.linear.default(mul_22, p_model_layers_0_mlp_gate_proj_weight); p_model_layers_0_mlp_gate_proj_weight = None
# File: ~/github/transformers/src/transformers/activations.py:103 in forward, code: return nn.functional.silu(input)
silu: "f32[s44, s70, 1024]" = torch.ops.aten.silu.default(linear_4); linear_4 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_5: "f32[s44, s70, 1024]" = torch.ops.aten.linear.default(mul_22, p_model_layers_0_mlp_up_proj_weight); mul_22 = p_model_layers_0_mlp_up_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:184 in forward, code: down_proj = self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x))
mul_23: "f32[s44, s70, 1024]" = torch.ops.aten.mul.Tensor(silu, linear_5); silu = linear_5 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_6: "f32[s44, s70, 192]" = torch.ops.aten.linear.default(mul_23, p_model_layers_0_mlp_down_proj_weight); mul_23 = p_model_layers_0_mlp_down_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:337 in forward, code: hidden_states = residual + hidden_states
add_9: "f32[s44, s70, 192]" = torch.ops.aten.add.Tensor(to_12, linear_6); to_12 = linear_6 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:64 in forward, code: hidden_states = hidden_states.to(torch.float32)
_assert_tensor_metadata_default_14 = torch.ops.aten._assert_tensor_metadata.default(add_9, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_14 = None
to_14: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(add_9, torch.float32); add_9 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:65 in forward, code: variance = hidden_states.pow(2).mean(-1, keepdim=True)
pow_3: "f32[s44, s70, 192]" = torch.ops.aten.pow.Tensor_Scalar(to_14, 2)
mean_2: "f32[s44, s70, 1]" = torch.ops.aten.mean.dim(pow_3, [-1], True); pow_3 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:66 in forward, code: hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
add_10: "f32[s44, s70, 1]" = torch.ops.aten.add.Tensor(mean_2, 1e-05); mean_2 = None
rsqrt_2: "f32[s44, s70, 1]" = torch.ops.aten.rsqrt.default(add_10); add_10 = None
mul_24: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(to_14, rsqrt_2); to_14 = rsqrt_2 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:67 in forward, code: return self.weight * hidden_states.to(input_dtype)
_assert_tensor_metadata_default_15 = torch.ops.aten._assert_tensor_metadata.default(mul_24, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_15 = None
to_15: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(mul_24, torch.float32); mul_24 = None
mul_25: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(p_model_norm_weight, to_15); p_model_norm_weight = to_15 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:500 in forward, code: logits = self.lm_head(hidden_states[:, slice_indices, :])
slice_12: "f32[s44, s70, 192]" = torch.ops.aten.slice.Tensor(mul_25, 0, 0, 9223372036854775807); mul_25 = None
slice_13: "f32[s44, s70, 192]" = torch.ops.aten.slice.Tensor(slice_12, 1, 0, 9223372036854775807); slice_12 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_7: "f32[s44, s70, 32000]" = torch.ops.aten.linear.default(slice_13, p_lm_head_weight); slice_13 = p_lm_head_weight = None
return (linear_7, cat_3, cat_4)
class submod_1(torch.nn.Module):
def forward(self, b_model_rotary_emb_inv_freq: "f32[48]", sym_size_int_19: "Sym(s44)", position_ids: "i64[s44, s70]"):
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:125 in forward, code: inv_freq_expanded = self.inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1).to(x.device)
unsqueeze: "f32[1, 48]" = torch.ops.aten.unsqueeze.default(b_model_rotary_emb_inv_freq, 0); b_model_rotary_emb_inv_freq = None
unsqueeze_1: "f32[1, 48, 1]" = torch.ops.aten.unsqueeze.default(unsqueeze, 2); unsqueeze = None
_assert_tensor_metadata_default_3 = torch.ops.aten._assert_tensor_metadata.default(unsqueeze_1, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_3 = None
to_3: "f32[1, 48, 1]" = torch.ops.aten.to.dtype(unsqueeze_1, torch.float32); unsqueeze_1 = None
expand_1: "f32[s44, 48, 1]" = torch.ops.aten.expand.default(to_3, [sym_size_int_19, -1, 1]); to_3 = sym_size_int_19 = None
_assert_tensor_metadata_default_4 = torch.ops.aten._assert_tensor_metadata.default(expand_1, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_4 = None
to_4: "f32[s44, 48, 1]" = torch.ops.aten.to.dtype_layout(expand_1, dtype = torch.float32, layout = torch.strided, device = device(type='cpu')); expand_1 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:126 in forward, code: position_ids_expanded = position_ids[:, None, :].float()
slice_1: "i64[s44, s70]" = torch.ops.aten.slice.Tensor(position_ids, 0, 0, 9223372036854775807); position_ids = None
unsqueeze_2: "i64[s44, 1, s70]" = torch.ops.aten.unsqueeze.default(slice_1, 1); slice_1 = None
slice_2: "i64[s44, 1, s70]" = torch.ops.aten.slice.Tensor(unsqueeze_2, 2, 0, 9223372036854775807); unsqueeze_2 = None
_assert_tensor_metadata_default_5 = torch.ops.aten._assert_tensor_metadata.default(slice_2, dtype = torch.int64, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_5 = None
to_5: "f32[s44, 1, s70]" = torch.ops.aten.to.dtype(slice_2, torch.float32); slice_2 = None
# No stacktrace found for following nodes
submod_3 = self.submod_1
wrap_with_autocast = torch.ops.higher_order.wrap_with_autocast('cpu', torch.bfloat16, False, False, submod_3, to_4, to_5); submod_3 = to_4 = to_5 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:132 in forward, code: cos = emb.cos() * self.attention_scaling
mul_5: "f32[s44, s70, 96]" = wrap_with_autocast[0]
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:133 in forward, code: sin = emb.sin() * self.attention_scaling
mul_6: "f32[s44, s70, 96]" = wrap_with_autocast[1]; wrap_with_autocast = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:135 in forward, code: return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype)
_assert_tensor_metadata_default_8 = torch.ops.aten._assert_tensor_metadata.default(mul_5, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_8 = None
to_8: "f32[s44, s70, 96]" = torch.ops.aten.to.dtype(mul_5, torch.float32); mul_5 = None
_assert_tensor_metadata_default_9 = torch.ops.aten._assert_tensor_metadata.default(mul_6, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_9 = None
to_9: "f32[s44, s70, 96]" = torch.ops.aten.to.dtype(mul_6, torch.float32); mul_6 = None
return (to_8, to_9)
class submod_1(torch.nn.Module):
def forward(self, to_4: "f32[s44, 48, 1]", to_5: "f32[s44, 1, s70]"):
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:130 in forward, code: freqs = (inv_freq_expanded.float() @ position_ids_expanded.float()).transpose(1, 2)
_assert_tensor_metadata_default_6 = torch.ops.aten._assert_tensor_metadata.default(to_4, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_6 = None
to_6: "f32[s44, 48, 1]" = torch.ops.aten.to.dtype(to_4, torch.float32); to_4 = None
_assert_tensor_metadata_default_7 = torch.ops.aten._assert_tensor_metadata.default(to_5, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_7 = None
to_7: "f32[s44, 1, s70]" = torch.ops.aten.to.dtype(to_5, torch.float32); to_5 = None
matmul: "f32[s44, 48, s70]" = torch.ops.aten.matmul.default(to_6, to_7); to_6 = to_7 = None
transpose: "f32[s44, s70, 48]" = torch.ops.aten.transpose.int(matmul, 1, 2); matmul = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:131 in forward, code: emb = torch.cat((freqs, freqs), dim=-1)
cat: "f32[s44, s70, 96]" = torch.ops.aten.cat.default([transpose, transpose], -1); transpose = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:132 in forward, code: cos = emb.cos() * self.attention_scaling
cos: "f32[s44, s70, 96]" = torch.ops.aten.cos.default(cat)
mul_5: "f32[s44, s70, 96]" = torch.ops.aten.mul.Tensor(cos, 1.0); cos = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:133 in forward, code: sin = emb.sin() * self.attention_scaling
sin: "f32[s44, s70, 96]" = torch.ops.aten.sin.default(cat); cat = None
mul_6: "f32[s44, s70, 96]" = torch.ops.aten.mul.Tensor(sin, 1.0); sin = None
return (mul_5, mul_6)
Graph signature:
# inputs
p_model_embed_tokens_weight: PARAMETER target='model.embed_tokens.weight'
p_model_layers_0_self_attn_q_proj_weight: PARAMETER target='model.layers.0.self_attn.q_proj.weight'
p_model_layers_0_self_attn_k_proj_weight: PARAMETER target='model.layers.0.self_attn.k_proj.weight'
p_model_layers_0_self_attn_v_proj_weight: PARAMETER target='model.layers.0.self_attn.v_proj.weight'
p_model_layers_0_self_attn_o_proj_weight: PARAMETER target='model.layers.0.self_attn.o_proj.weight'
p_model_layers_0_mlp_gate_proj_weight: PARAMETER target='model.layers.0.mlp.gate_proj.weight'
p_model_layers_0_mlp_up_proj_weight: PARAMETER target='model.layers.0.mlp.up_proj.weight'
p_model_layers_0_mlp_down_proj_weight: PARAMETER target='model.layers.0.mlp.down_proj.weight'
p_model_layers_0_input_layernorm_weight: PARAMETER target='model.layers.0.input_layernorm.weight'
p_model_layers_0_post_attention_layernorm_weight: PARAMETER target='model.layers.0.post_attention_layernorm.weight'
p_model_norm_weight: PARAMETER target='model.norm.weight'
p_lm_head_weight: PARAMETER target='lm_head.weight'
b_model_rotary_emb_inv_freq: BUFFER target='model.rotary_emb.inv_freq' persistent=False
input_ids: USER_INPUT
attention_mask: USER_INPUT
position_ids: USER_INPUT
past_key_values_key_0: USER_INPUT
past_key_values_value_0: USER_INPUT
# outputs
linear_7: USER_OUTPUT
cat_3: USER_OUTPUT
cat_4: USER_OUTPUT
Range constraints: {s44: VR[0, int_oo], s70: VR[0, int_oo], s43: VR[0, int_oo], s53: VR[0, int_oo], s45: VR[0, int_oo], s21: VR[0, int_oo]}
Back to the original model¶
Let’s use the same dummy inputs but we use the downloaded model.
Dummy inputs and dynamic shapes are created by function
onnx_diagnostic.torch_models.llms.get_tiny_llm().
data = get_tiny_llm()
inputs, dynamic_shapes = data["inputs"], data["dynamic_shapes"]
Let’s print the inputs.
print(string_type(inputs, with_shape=True))
dict(input_ids:T7s2x3,attention_mask:T7s2x33,position_ids:T7s2x3,past_key_values:DynamicCache(key_cache=#1[T1s2x1x30x96], value_cache=#1[T1s2x1x30x96]))
{'attention_mask': {0: 'batch', 1: 'cache+seq'},
'input_ids': {0: 'batch', 1: 'seq_length'},
'past_key_values': [{0: 'batch', 2: 'cache_length'},
{0: 'batch', 2: 'cache_length'}],
'position_ids': {0: 'batch', 1: 'seq_length'}}
And Let’s finally export.
try:
ep = torch.export.export(
model,
(),
kwargs=cloned_inputs,
dynamic_shapes=use_dyn_not_str(dynamic_shapes),
strict=False,
)
print("It worked:")
print(ep)
except Exception as e:
# To work, it needs at least PRs:
# * https://github.com/huggingface/transformers/pull/36311
# * https://github.com/huggingface/transformers/pull/36652
print("It failed:", e)
It worked:
ExportedProgram:
class GraphModule(torch.nn.Module):
def forward(self, p_model_embed_tokens_weight: "f32[32000, 192]", p_model_layers_0_self_attn_q_proj_weight: "f32[192, 192]", p_model_layers_0_self_attn_k_proj_weight: "f32[96, 192]", p_model_layers_0_self_attn_v_proj_weight: "f32[96, 192]", p_model_layers_0_self_attn_o_proj_weight: "f32[192, 192]", p_model_layers_0_mlp_gate_proj_weight: "f32[1024, 192]", p_model_layers_0_mlp_up_proj_weight: "f32[1024, 192]", p_model_layers_0_mlp_down_proj_weight: "f32[192, 1024]", p_model_layers_0_input_layernorm_weight: "f32[192]", p_model_layers_0_post_attention_layernorm_weight: "f32[192]", p_model_norm_weight: "f32[192]", p_lm_head_weight: "f32[32000, 192]", b_model_rotary_emb_inv_freq: "f32[48]", input_ids: "i64[s44, s70]", attention_mask: "i64[s43, s53]", position_ids: "i64[s44, s70]", past_key_values_key_0: "f32[s44, 1, s45, 96]", past_key_values_value_0: "f32[s44, 1, s21, 96]"):
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:409 in forward, code: causal_mask = create_causal_mask(
function_const_func_spec0 = self.function_const_func_spec0
torch__dynamo__trace_wrapped_higher_order_op_mod_index0 = self.torch__dynamo__trace_wrapped_higher_order_op_ModIndex0
#
sym_size_int_16: "Sym(s70)" = torch.ops.aten.sym_size.int(input_ids, 1)
sym_size_int_19: "Sym(s44)" = torch.ops.aten.sym_size.int(position_ids, 0)
sym_size_int_22: "Sym(s45)" = torch.ops.aten.sym_size.int(past_key_values_key_0, 2)
sym_size_int_24: "Sym(s21)" = torch.ops.aten.sym_size.int(past_key_values_value_0, 2)
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/sparse.py:192 in forward, code: return F.embedding(
embedding: "f32[s44, s70, 192]" = torch.ops.aten.embedding.default(p_model_embed_tokens_weight, input_ids); p_model_embed_tokens_weight = input_ids = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:403 in forward, code: past_seen_tokens, past_seen_tokens + inputs_embeds.shape[1], device=inputs_embeds.device
add: "Sym(s45 + s70)" = sym_size_int_22 + sym_size_int_16
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:402 in forward, code: cache_position: torch.Tensor = torch.arange(
arange: "i64[s70]" = torch.ops.aten.arange.start(sym_size_int_22, add, device = device(type='cpu'), pin_memory = False); sym_size_int_22 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:409 in forward, code: causal_mask = create_causal_mask(
_assert_tensor_metadata_default = torch.ops.aten._assert_tensor_metadata.default(attention_mask, dtype = torch.int64, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default = None
to: "b8[s43, s53]" = torch.ops.aten.to.device(attention_mask, device(type='cpu'), torch.bool); attention_mask = None
arange_1: "i64[s45 + s70]" = torch.ops.aten.arange.default(add, device = device(type='cpu'), pin_memory = False)
add_: "i64[s45 + s70]" = torch.ops.aten.add_.Tensor(arange_1, 0); arange_1 = None
arange_2: "i64[s44]" = torch.ops.aten.arange.default(sym_size_int_19, device = device(type='cpu'), pin_memory = False)
arange_3: "i64[1]" = torch.ops.aten.arange.default(1, device = device(type='cpu'), pin_memory = False)
lazy_load_decompositions = torch._functorch.predispatch.lazy_load_decompositions(); lazy_load_decompositions = None
_vmap_increment_nesting = torch._functorch.predispatch._vmap_increment_nesting(sym_size_int_19, 'error'); _vmap_increment_nesting = None
_add_batch_dim: "i64[]" = torch._functorch.predispatch._add_batch_dim(arange_2, 0, 1); arange_2 = None
lazy_load_decompositions_1 = torch._functorch.predispatch.lazy_load_decompositions(); lazy_load_decompositions_1 = None
_vmap_increment_nesting_1 = torch._functorch.predispatch._vmap_increment_nesting(1, 'error'); _vmap_increment_nesting_1 = None
_add_batch_dim_1: "i64[]" = torch._functorch.predispatch._add_batch_dim(arange_3, 0, 2); arange_3 = _add_batch_dim_1 = None
lazy_load_decompositions_2 = torch._functorch.predispatch.lazy_load_decompositions(); lazy_load_decompositions_2 = None
_vmap_increment_nesting_2 = torch._functorch.predispatch._vmap_increment_nesting(sym_size_int_16, 'error'); _vmap_increment_nesting_2 = None
_add_batch_dim_2: "i64[]" = torch._functorch.predispatch._add_batch_dim(arange, 0, 3); arange = None
lazy_load_decompositions_3 = torch._functorch.predispatch.lazy_load_decompositions(); lazy_load_decompositions_3 = None
_vmap_increment_nesting_3 = torch._functorch.predispatch._vmap_increment_nesting(add, 'error'); _vmap_increment_nesting_3 = None
_add_batch_dim_3: "i64[]" = torch._functorch.predispatch._add_batch_dim(add_, 0, 4); add_ = None
new_ones: "b8[]" = torch.ops.aten.new_ones.default(_add_batch_dim_2, [], dtype = torch.bool, pin_memory = False)
le: "b8[]" = torch.ops.aten.le.Tensor(_add_batch_dim_3, _add_batch_dim_2); _add_batch_dim_2 = None
_assert_tensor_metadata_default_1 = torch.ops.aten._assert_tensor_metadata.default(le, dtype = torch.bool, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_1 = None
to_1: "b8[]" = torch.ops.aten.to.dtype_layout(le, dtype = torch.bool, layout = torch.strided, device = device(type='cpu')); le = None
and_1: "b8[]" = torch.ops.aten.__and__.Tensor(new_ones, to_1); new_ones = to_1 = None
flat_apply: "b8[]" = torch.ops.higher_order.flat_apply(function_const_func_spec0, torch__dynamo__trace_wrapped_higher_order_op_mod_index0, 'torch._dynamo._trace_wrapped_higher_order_op.ModIndex', to, _add_batch_dim, _add_batch_dim_3); function_const_func_spec0 = torch__dynamo__trace_wrapped_higher_order_op_mod_index0 = to = _add_batch_dim = _add_batch_dim_3 = None
_assert_tensor_metadata_default_2 = torch.ops.aten._assert_tensor_metadata.default(flat_apply, dtype = torch.bool, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_2 = None
to_2: "b8[]" = torch.ops.aten.to.dtype_layout(flat_apply, dtype = torch.bool, layout = torch.strided, device = device(type='cpu')); flat_apply = None
and_2: "b8[]" = torch.ops.aten.__and__.Tensor(and_1, to_2); and_1 = to_2 = None
_remove_batch_dim: "b8[s45 + s70]" = torch._functorch.predispatch._remove_batch_dim(and_2, 4, add, 0); and_2 = None
_vmap_decrement_nesting = torch._functorch.predispatch._vmap_decrement_nesting(); _vmap_decrement_nesting = None
_remove_batch_dim_1: "b8[s70, s45 + s70]" = torch._functorch.predispatch._remove_batch_dim(_remove_batch_dim, 3, sym_size_int_16, 0); _remove_batch_dim = None
_vmap_decrement_nesting_1 = torch._functorch.predispatch._vmap_decrement_nesting(); _vmap_decrement_nesting_1 = None
_remove_batch_dim_2: "b8[1, s70, s45 + s70]" = torch._functorch.predispatch._remove_batch_dim(_remove_batch_dim_1, 2, 1, 0)
expand: "b8[1, s70, s45 + s70]" = torch.ops.aten.expand.default(_remove_batch_dim_1, [1, sym_size_int_16, add]); _remove_batch_dim_1 = expand = None
_vmap_decrement_nesting_2 = torch._functorch.predispatch._vmap_decrement_nesting(); _vmap_decrement_nesting_2 = None
_remove_batch_dim_3: "b8[s44, 1, s70, s45 + s70]" = torch._functorch.predispatch._remove_batch_dim(_remove_batch_dim_2, 1, sym_size_int_19, 0); _remove_batch_dim_2 = None
_vmap_decrement_nesting_3 = torch._functorch.predispatch._vmap_decrement_nesting(); _vmap_decrement_nesting_3 = None
# No stacktrace found for following nodes
submod_3 = self.submod_1
wrap_with_set_grad_enabled = torch.ops.higher_order.wrap_with_set_grad_enabled(False, submod_3, b_model_rotary_emb_inv_freq, sym_size_int_19, position_ids); submod_3 = b_model_rotary_emb_inv_freq = position_ids = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:135 in forward, code: return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype)
to_8: "f32[s44, s70, 96]" = wrap_with_set_grad_enabled[0]
to_9: "f32[s44, s70, 96]" = wrap_with_set_grad_enabled[1]; wrap_with_set_grad_enabled = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:64 in forward, code: hidden_states = hidden_states.to(torch.float32)
_assert_tensor_metadata_default_10 = torch.ops.aten._assert_tensor_metadata.default(embedding, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_10 = None
to_10: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(embedding, torch.float32); embedding = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:65 in forward, code: variance = hidden_states.pow(2).mean(-1, keepdim=True)
pow_1: "f32[s44, s70, 192]" = torch.ops.aten.pow.Tensor_Scalar(to_10, 2)
mean: "f32[s44, s70, 1]" = torch.ops.aten.mean.dim(pow_1, [-1], True); pow_1 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:66 in forward, code: hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
add_4: "f32[s44, s70, 1]" = torch.ops.aten.add.Tensor(mean, 1e-05); mean = None
rsqrt: "f32[s44, s70, 1]" = torch.ops.aten.rsqrt.default(add_4); add_4 = None
mul_7: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(to_10, rsqrt); rsqrt = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:67 in forward, code: return self.weight * hidden_states.to(input_dtype)
_assert_tensor_metadata_default_11 = torch.ops.aten._assert_tensor_metadata.default(mul_7, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_11 = None
to_11: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(mul_7, torch.float32); mul_7 = None
mul_8: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(p_model_layers_0_input_layernorm_weight, to_11); p_model_layers_0_input_layernorm_weight = to_11 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear: "f32[s44, s70, 192]" = torch.ops.aten.linear.default(mul_8, p_model_layers_0_self_attn_q_proj_weight); p_model_layers_0_self_attn_q_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:264 in forward, code: query_states = self.q_proj(hidden_states).view(hidden_shape).transpose(1, 2)
view: "f32[s44, s70, 2, 96]" = torch.ops.aten.view.default(linear, [sym_size_int_19, sym_size_int_16, -1, 96]); linear = None
transpose_1: "f32[s44, 2, s70, 96]" = torch.ops.aten.transpose.int(view, 1, 2); view = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_1: "f32[s44, s70, 96]" = torch.ops.aten.linear.default(mul_8, p_model_layers_0_self_attn_k_proj_weight); p_model_layers_0_self_attn_k_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:265 in forward, code: key_states = self.k_proj(hidden_states).view(hidden_shape).transpose(1, 2)
view_1: "f32[s44, s70, 1, 96]" = torch.ops.aten.view.default(linear_1, [sym_size_int_19, sym_size_int_16, -1, 96]); linear_1 = None
transpose_2: "f32[s44, 1, s70, 96]" = torch.ops.aten.transpose.int(view_1, 1, 2); view_1 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_2: "f32[s44, s70, 96]" = torch.ops.aten.linear.default(mul_8, p_model_layers_0_self_attn_v_proj_weight); mul_8 = p_model_layers_0_self_attn_v_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:266 in forward, code: value_states = self.v_proj(hidden_states).view(hidden_shape).transpose(1, 2)
view_2: "f32[s44, s70, 1, 96]" = torch.ops.aten.view.default(linear_2, [sym_size_int_19, sym_size_int_16, -1, 96]); linear_2 = None
transpose_3: "f32[s44, 1, s70, 96]" = torch.ops.aten.transpose.int(view_2, 1, 2); view_2 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:269 in forward, code: query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin)
unsqueeze_3: "f32[s44, 1, s70, 96]" = torch.ops.aten.unsqueeze.default(to_8, 1); to_8 = None
unsqueeze_4: "f32[s44, 1, s70, 96]" = torch.ops.aten.unsqueeze.default(to_9, 1); to_9 = None
mul_9: "f32[s44, 2, s70, 96]" = torch.ops.aten.mul.Tensor(transpose_1, unsqueeze_3)
slice_3: "f32[s44, 2, s70, 48]" = torch.ops.aten.slice.Tensor(transpose_1, 3, 0, 48)
slice_4: "f32[s44, 2, s70, 48]" = torch.ops.aten.slice.Tensor(transpose_1, 3, 48, 9223372036854775807); transpose_1 = None
neg: "f32[s44, 2, s70, 48]" = torch.ops.aten.neg.default(slice_4); slice_4 = None
cat_1: "f32[s44, 2, s70, 96]" = torch.ops.aten.cat.default([neg, slice_3], -1); neg = slice_3 = None
mul_10: "f32[s44, 2, s70, 96]" = torch.ops.aten.mul.Tensor(cat_1, unsqueeze_4); cat_1 = None
add_5: "f32[s44, 2, s70, 96]" = torch.ops.aten.add.Tensor(mul_9, mul_10); mul_9 = mul_10 = None
mul_11: "f32[s44, 1, s70, 96]" = torch.ops.aten.mul.Tensor(transpose_2, unsqueeze_3); unsqueeze_3 = None
slice_5: "f32[s44, 1, s70, 48]" = torch.ops.aten.slice.Tensor(transpose_2, 3, 0, 48)
slice_6: "f32[s44, 1, s70, 48]" = torch.ops.aten.slice.Tensor(transpose_2, 3, 48, 9223372036854775807); transpose_2 = None
neg_1: "f32[s44, 1, s70, 48]" = torch.ops.aten.neg.default(slice_6); slice_6 = None
cat_2: "f32[s44, 1, s70, 96]" = torch.ops.aten.cat.default([neg_1, slice_5], -1); neg_1 = slice_5 = None
mul_12: "f32[s44, 1, s70, 96]" = torch.ops.aten.mul.Tensor(cat_2, unsqueeze_4); cat_2 = unsqueeze_4 = None
add_6: "f32[s44, 1, s70, 96]" = torch.ops.aten.add.Tensor(mul_11, mul_12); mul_11 = mul_12 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:274 in forward, code: key_states, value_states = past_key_values.update(key_states, value_states, self.layer_idx, cache_kwargs)
cat_3: "f32[s44, 1, s45 + s70, 96]" = torch.ops.aten.cat.default([past_key_values_key_0, add_6], -2); past_key_values_key_0 = add_6 = None
cat_4: "f32[s44, 1, s21 + s70, 96]" = torch.ops.aten.cat.default([past_key_values_value_0, transpose_3], -2); past_key_values_value_0 = transpose_3 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:280 in forward, code: attn_output, attn_weights = attention_interface(
slice_7: "f32[s44, 1, s45 + s70, 96]" = torch.ops.aten.slice.Tensor(cat_3, 0, 0, 9223372036854775807)
unsqueeze_5: "f32[s44, 1, 1, s45 + s70, 96]" = torch.ops.aten.unsqueeze.default(slice_7, 2); slice_7 = None
slice_8: "f32[s44, 1, 1, s45 + s70, 96]" = torch.ops.aten.slice.Tensor(unsqueeze_5, 3, 0, 9223372036854775807); unsqueeze_5 = None
expand_2: "f32[s44, 1, 2, s45 + s70, 96]" = torch.ops.aten.expand.default(slice_8, [sym_size_int_19, 1, 2, add, 96]); slice_8 = None
reshape: "f32[s44, 2, s45 + s70, 96]" = torch.ops.aten.reshape.default(expand_2, [sym_size_int_19, 2, add, 96]); expand_2 = None
slice_9: "f32[s44, 1, s21 + s70, 96]" = torch.ops.aten.slice.Tensor(cat_4, 0, 0, 9223372036854775807)
unsqueeze_6: "f32[s44, 1, 1, s21 + s70, 96]" = torch.ops.aten.unsqueeze.default(slice_9, 2); slice_9 = None
add_11: "Sym(s21 + s70)" = sym_size_int_24 + sym_size_int_16; sym_size_int_24 = None
slice_10: "f32[s44, 1, 1, s21 + s70, 96]" = torch.ops.aten.slice.Tensor(unsqueeze_6, 3, 0, 9223372036854775807); unsqueeze_6 = None
expand_3: "f32[s44, 1, 2, s21 + s70, 96]" = torch.ops.aten.expand.default(slice_10, [sym_size_int_19, 1, 2, add_11, 96]); slice_10 = None
reshape_1: "f32[s44, 2, s21 + s70, 96]" = torch.ops.aten.reshape.default(expand_3, [sym_size_int_19, 2, add_11, 96]); expand_3 = add_11 = None
slice_11: "b8[s44, 1, s70, s45 + s70]" = torch.ops.aten.slice.Tensor(_remove_batch_dim_3, 3, None, add); _remove_batch_dim_3 = add = None
scaled_dot_product_attention: "f32[s44, 2, s70, 96]" = torch.ops.aten.scaled_dot_product_attention.default(add_5, reshape, reshape_1, slice_11, scale = 0.10206207261596575); add_5 = reshape = reshape_1 = slice_11 = None
transpose_4: "f32[s44, s70, 2, 96]" = torch.ops.aten.transpose.int(scaled_dot_product_attention, 1, 2); scaled_dot_product_attention = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:291 in forward, code: attn_output = attn_output.reshape(*input_shape, -1).contiguous()
reshape_2: "f32[s44, s70, 192]" = torch.ops.aten.reshape.default(transpose_4, [sym_size_int_19, sym_size_int_16, -1]); transpose_4 = sym_size_int_19 = sym_size_int_16 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_3: "f32[s44, s70, 192]" = torch.ops.aten.linear.default(reshape_2, p_model_layers_0_self_attn_o_proj_weight); reshape_2 = p_model_layers_0_self_attn_o_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:331 in forward, code: hidden_states = residual + hidden_states
add_7: "f32[s44, s70, 192]" = torch.ops.aten.add.Tensor(to_10, linear_3); to_10 = linear_3 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:64 in forward, code: hidden_states = hidden_states.to(torch.float32)
_assert_tensor_metadata_default_12 = torch.ops.aten._assert_tensor_metadata.default(add_7, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_12 = None
to_12: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(add_7, torch.float32); add_7 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:65 in forward, code: variance = hidden_states.pow(2).mean(-1, keepdim=True)
pow_2: "f32[s44, s70, 192]" = torch.ops.aten.pow.Tensor_Scalar(to_12, 2)
mean_1: "f32[s44, s70, 1]" = torch.ops.aten.mean.dim(pow_2, [-1], True); pow_2 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:66 in forward, code: hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
add_8: "f32[s44, s70, 1]" = torch.ops.aten.add.Tensor(mean_1, 1e-05); mean_1 = None
rsqrt_1: "f32[s44, s70, 1]" = torch.ops.aten.rsqrt.default(add_8); add_8 = None
mul_21: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(to_12, rsqrt_1); rsqrt_1 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:67 in forward, code: return self.weight * hidden_states.to(input_dtype)
_assert_tensor_metadata_default_13 = torch.ops.aten._assert_tensor_metadata.default(mul_21, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_13 = None
to_13: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(mul_21, torch.float32); mul_21 = None
mul_22: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(p_model_layers_0_post_attention_layernorm_weight, to_13); p_model_layers_0_post_attention_layernorm_weight = to_13 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_4: "f32[s44, s70, 1024]" = torch.ops.aten.linear.default(mul_22, p_model_layers_0_mlp_gate_proj_weight); p_model_layers_0_mlp_gate_proj_weight = None
# File: ~/github/transformers/src/transformers/activations.py:103 in forward, code: return nn.functional.silu(input)
silu: "f32[s44, s70, 1024]" = torch.ops.aten.silu.default(linear_4); linear_4 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_5: "f32[s44, s70, 1024]" = torch.ops.aten.linear.default(mul_22, p_model_layers_0_mlp_up_proj_weight); mul_22 = p_model_layers_0_mlp_up_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:184 in forward, code: down_proj = self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x))
mul_23: "f32[s44, s70, 1024]" = torch.ops.aten.mul.Tensor(silu, linear_5); silu = linear_5 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_6: "f32[s44, s70, 192]" = torch.ops.aten.linear.default(mul_23, p_model_layers_0_mlp_down_proj_weight); mul_23 = p_model_layers_0_mlp_down_proj_weight = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:337 in forward, code: hidden_states = residual + hidden_states
add_9: "f32[s44, s70, 192]" = torch.ops.aten.add.Tensor(to_12, linear_6); to_12 = linear_6 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:64 in forward, code: hidden_states = hidden_states.to(torch.float32)
_assert_tensor_metadata_default_14 = torch.ops.aten._assert_tensor_metadata.default(add_9, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_14 = None
to_14: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(add_9, torch.float32); add_9 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:65 in forward, code: variance = hidden_states.pow(2).mean(-1, keepdim=True)
pow_3: "f32[s44, s70, 192]" = torch.ops.aten.pow.Tensor_Scalar(to_14, 2)
mean_2: "f32[s44, s70, 1]" = torch.ops.aten.mean.dim(pow_3, [-1], True); pow_3 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:66 in forward, code: hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
add_10: "f32[s44, s70, 1]" = torch.ops.aten.add.Tensor(mean_2, 1e-05); mean_2 = None
rsqrt_2: "f32[s44, s70, 1]" = torch.ops.aten.rsqrt.default(add_10); add_10 = None
mul_24: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(to_14, rsqrt_2); to_14 = rsqrt_2 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:67 in forward, code: return self.weight * hidden_states.to(input_dtype)
_assert_tensor_metadata_default_15 = torch.ops.aten._assert_tensor_metadata.default(mul_24, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_15 = None
to_15: "f32[s44, s70, 192]" = torch.ops.aten.to.dtype(mul_24, torch.float32); mul_24 = None
mul_25: "f32[s44, s70, 192]" = torch.ops.aten.mul.Tensor(p_model_norm_weight, to_15); p_model_norm_weight = to_15 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:500 in forward, code: logits = self.lm_head(hidden_states[:, slice_indices, :])
slice_12: "f32[s44, s70, 192]" = torch.ops.aten.slice.Tensor(mul_25, 0, 0, 9223372036854775807); mul_25 = None
slice_13: "f32[s44, s70, 192]" = torch.ops.aten.slice.Tensor(slice_12, 1, 0, 9223372036854775807); slice_12 = None
# File: ~/vv/this312/lib/python3.12/site-packages/torch/nn/modules/linear.py:134 in forward, code: return F.linear(input, self.weight, self.bias)
linear_7: "f32[s44, s70, 32000]" = torch.ops.aten.linear.default(slice_13, p_lm_head_weight); slice_13 = p_lm_head_weight = None
return (linear_7, cat_3, cat_4)
class submod_1(torch.nn.Module):
def forward(self, b_model_rotary_emb_inv_freq: "f32[48]", sym_size_int_19: "Sym(s44)", position_ids: "i64[s44, s70]"):
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:125 in forward, code: inv_freq_expanded = self.inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1).to(x.device)
unsqueeze: "f32[1, 48]" = torch.ops.aten.unsqueeze.default(b_model_rotary_emb_inv_freq, 0); b_model_rotary_emb_inv_freq = None
unsqueeze_1: "f32[1, 48, 1]" = torch.ops.aten.unsqueeze.default(unsqueeze, 2); unsqueeze = None
_assert_tensor_metadata_default_3 = torch.ops.aten._assert_tensor_metadata.default(unsqueeze_1, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_3 = None
to_3: "f32[1, 48, 1]" = torch.ops.aten.to.dtype(unsqueeze_1, torch.float32); unsqueeze_1 = None
expand_1: "f32[s44, 48, 1]" = torch.ops.aten.expand.default(to_3, [sym_size_int_19, -1, 1]); to_3 = sym_size_int_19 = None
_assert_tensor_metadata_default_4 = torch.ops.aten._assert_tensor_metadata.default(expand_1, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_4 = None
to_4: "f32[s44, 48, 1]" = torch.ops.aten.to.dtype_layout(expand_1, dtype = torch.float32, layout = torch.strided, device = device(type='cpu')); expand_1 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:126 in forward, code: position_ids_expanded = position_ids[:, None, :].float()
slice_1: "i64[s44, s70]" = torch.ops.aten.slice.Tensor(position_ids, 0, 0, 9223372036854775807); position_ids = None
unsqueeze_2: "i64[s44, 1, s70]" = torch.ops.aten.unsqueeze.default(slice_1, 1); slice_1 = None
slice_2: "i64[s44, 1, s70]" = torch.ops.aten.slice.Tensor(unsqueeze_2, 2, 0, 9223372036854775807); unsqueeze_2 = None
_assert_tensor_metadata_default_5 = torch.ops.aten._assert_tensor_metadata.default(slice_2, dtype = torch.int64, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_5 = None
to_5: "f32[s44, 1, s70]" = torch.ops.aten.to.dtype(slice_2, torch.float32); slice_2 = None
# No stacktrace found for following nodes
submod_3 = self.submod_1
wrap_with_autocast = torch.ops.higher_order.wrap_with_autocast('cpu', torch.bfloat16, False, False, submod_3, to_4, to_5); submod_3 = to_4 = to_5 = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:132 in forward, code: cos = emb.cos() * self.attention_scaling
mul_5: "f32[s44, s70, 96]" = wrap_with_autocast[0]
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:133 in forward, code: sin = emb.sin() * self.attention_scaling
mul_6: "f32[s44, s70, 96]" = wrap_with_autocast[1]; wrap_with_autocast = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:135 in forward, code: return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype)
_assert_tensor_metadata_default_8 = torch.ops.aten._assert_tensor_metadata.default(mul_5, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_8 = None
to_8: "f32[s44, s70, 96]" = torch.ops.aten.to.dtype(mul_5, torch.float32); mul_5 = None
_assert_tensor_metadata_default_9 = torch.ops.aten._assert_tensor_metadata.default(mul_6, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_9 = None
to_9: "f32[s44, s70, 96]" = torch.ops.aten.to.dtype(mul_6, torch.float32); mul_6 = None
return (to_8, to_9)
class submod_1(torch.nn.Module):
def forward(self, to_4: "f32[s44, 48, 1]", to_5: "f32[s44, 1, s70]"):
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:130 in forward, code: freqs = (inv_freq_expanded.float() @ position_ids_expanded.float()).transpose(1, 2)
_assert_tensor_metadata_default_6 = torch.ops.aten._assert_tensor_metadata.default(to_4, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_6 = None
to_6: "f32[s44, 48, 1]" = torch.ops.aten.to.dtype(to_4, torch.float32); to_4 = None
_assert_tensor_metadata_default_7 = torch.ops.aten._assert_tensor_metadata.default(to_5, dtype = torch.float32, device = device(type='cpu'), layout = torch.strided); _assert_tensor_metadata_default_7 = None
to_7: "f32[s44, 1, s70]" = torch.ops.aten.to.dtype(to_5, torch.float32); to_5 = None
matmul: "f32[s44, 48, s70]" = torch.ops.aten.matmul.default(to_6, to_7); to_6 = to_7 = None
transpose: "f32[s44, s70, 48]" = torch.ops.aten.transpose.int(matmul, 1, 2); matmul = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:131 in forward, code: emb = torch.cat((freqs, freqs), dim=-1)
cat: "f32[s44, s70, 96]" = torch.ops.aten.cat.default([transpose, transpose], -1); transpose = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:132 in forward, code: cos = emb.cos() * self.attention_scaling
cos: "f32[s44, s70, 96]" = torch.ops.aten.cos.default(cat)
mul_5: "f32[s44, s70, 96]" = torch.ops.aten.mul.Tensor(cos, 1.0); cos = None
# File: ~/github/transformers/src/transformers/models/llama/modeling_llama.py:133 in forward, code: sin = emb.sin() * self.attention_scaling
sin: "f32[s44, s70, 96]" = torch.ops.aten.sin.default(cat); cat = None
mul_6: "f32[s44, s70, 96]" = torch.ops.aten.mul.Tensor(sin, 1.0); sin = None
return (mul_5, mul_6)
Graph signature:
# inputs
p_model_embed_tokens_weight: PARAMETER target='model.embed_tokens.weight'
p_model_layers_0_self_attn_q_proj_weight: PARAMETER target='model.layers.0.self_attn.q_proj.weight'
p_model_layers_0_self_attn_k_proj_weight: PARAMETER target='model.layers.0.self_attn.k_proj.weight'
p_model_layers_0_self_attn_v_proj_weight: PARAMETER target='model.layers.0.self_attn.v_proj.weight'
p_model_layers_0_self_attn_o_proj_weight: PARAMETER target='model.layers.0.self_attn.o_proj.weight'
p_model_layers_0_mlp_gate_proj_weight: PARAMETER target='model.layers.0.mlp.gate_proj.weight'
p_model_layers_0_mlp_up_proj_weight: PARAMETER target='model.layers.0.mlp.up_proj.weight'
p_model_layers_0_mlp_down_proj_weight: PARAMETER target='model.layers.0.mlp.down_proj.weight'
p_model_layers_0_input_layernorm_weight: PARAMETER target='model.layers.0.input_layernorm.weight'
p_model_layers_0_post_attention_layernorm_weight: PARAMETER target='model.layers.0.post_attention_layernorm.weight'
p_model_norm_weight: PARAMETER target='model.norm.weight'
p_lm_head_weight: PARAMETER target='lm_head.weight'
b_model_rotary_emb_inv_freq: BUFFER target='model.rotary_emb.inv_freq' persistent=False
input_ids: USER_INPUT
attention_mask: USER_INPUT
position_ids: USER_INPUT
past_key_values_key_0: USER_INPUT
past_key_values_value_0: USER_INPUT
# outputs
linear_7: USER_OUTPUT
cat_3: USER_OUTPUT
cat_4: USER_OUTPUT
Range constraints: {s44: VR[0, int_oo], s70: VR[0, int_oo], s43: VR[0, int_oo], s53: VR[0, int_oo], s45: VR[0, int_oo], s21: VR[0, int_oo]}
If you have any error, then look at example Export Tiny-LLM with patches.
doc.plot_legend("Tiny-LLM\nforward inputs\nbehind generate", "torch.export.export", "tomato")
Total running time of the script: (0 minutes 5.078 seconds)
Related examples
Export with DynamicCache and guessed dynamic shapes
Export with DynamicCache and guessed dynamic shapes