Test

.gitignore

@@ -0,0 +1,2 @@
+.venv/
+.cache/

README.md

@@ -1,3 +1,11 @@
 # TFM
 
 Test TFM remote execution on upc server
+
+## TODO:
+
+- Alguna cosa legal per carregar models i fer benchmarking potser
+- Mirar de com implementar la cosa de busqueda de tensors
+- Reemplaçarlos allà
+- Fer posttraining
+- Profit

benchmark.sh

@@ -0,0 +1,6 @@
+#!/bin/bash
+lm_eval
+    --model hf \
+    --model_args pretrained=~/Uni/tfm/qwen_random,dtype=bfloat16,device_mp=auto \
+    --tasks hellaswag,arc_easy,boolq \
+    --batch_size auto

requirements.txt

@@ -0,0 +1,101 @@
+absl-py==2.4.0
+accelerate==1.13.0
+aiohappyeyeballs==2.6.2
+aiohttp==3.14.1
+aiosignal==1.4.0
+annotated-doc==0.0.4
+anyio==4.13.0
+attrs==26.1.0
+certifi==2026.5.20
+chardet==6.0.0.post1
+charset-normalizer==3.4.7
+click==8.4.1
+colorama==0.4.6
+cuda-bindings==13.3.1
+cuda-pathfinder==1.5.5
+cuda-toolkit==13.0.2
+DataProperty==1.1.1
+datasets==5.0.0
+dill==0.4.1
+evaluate==0.4.6
+filelock==3.29.1
+frozenlist==1.8.0
+fsspec==2026.4.0
+h11==0.16.0
+hf-xet==1.5.1
+httpcore==1.0.9
+httpx==0.28.1
+huggingface_hub==1.18.0
+idna==3.18
+Jinja2==3.1.6
+joblib==1.5.3
+lm_eval==0.4.12
+lxml==6.1.1
+markdown-it-py==4.2.0
+MarkupSafe==3.0.3
+mbstrdecoder==1.1.5
+mdurl==0.1.2
+more-itertools==11.1.0
+mpmath==1.3.0
+multidict==6.7.1
+multiprocess==0.70.19
+narwhals==2.22.1
+networkx==3.6.1
+nltk==3.9.4
+numpy==2.4.6
+nvidia-cublas==13.1.1.3
+nvidia-cuda-cupti==13.0.85
+nvidia-cuda-nvrtc==13.0.88
+nvidia-cuda-runtime==13.0.96
+nvidia-cudnn-cu13==9.20.0.48
+nvidia-cufft==12.0.0.61
+nvidia-cufile==1.15.1.6
+nvidia-curand==10.4.0.35
+nvidia-cusolver==12.0.4.66
+nvidia-cusparse==12.6.3.3
+nvidia-cusparselt-cu13==0.8.1
+nvidia-nccl-cu13==2.29.7
+nvidia-nvjitlink==13.0.88
+nvidia-nvshmem-cu13==3.4.5
+nvidia-nvtx==13.0.85
+packaging==26.2
+pandas==3.0.3
+pathvalidate==3.3.1
+portalocker==3.2.0
+propcache==0.5.2
+psutil==7.2.2
+pyarrow==24.0.0
+Pygments==2.20.0
+pytablewriter==1.2.1
+python-dateutil==2.9.0.post0
+pytz==2026.2
+PyYAML==6.0.3
+regex==2026.5.9
+requests==2.34.2
+rich==15.0.0
+rouge_score==0.1.2
+sacrebleu==2.6.0
+safetensors==0.8.0
+scikit-learn==1.9.0
+scipy==1.17.1
+setuptools==81.0.0
+shellingham==1.5.4
+six==1.17.0
+sqlitedict==2.1.0
+sympy==1.14.0
+tabledata==1.3.5
+tabulate==0.10.0
+tcolorpy==0.1.7
+threadpoolctl==3.6.0
+tokenizers==0.22.2
+torch==2.12.0
+tqdm==4.68.2
+transformers==5.10.2
+triton==3.7.0
+typepy==1.3.5
+typer==0.25.1
+typing_extensions==4.15.0
+urllib3==2.7.0
+word2number==1.1
+xxhash==3.7.0
+yarl==1.24.2

test.py

@@ -0,0 +1,65 @@
+import torch
+import torch.nn as nn
+from transformers import AutoTokenizer, AutoModelForCausalLM
+
+model_id = "Qwen/Qwen3.5-0.8B"
+cache_path = "./.cache"
+
+# Load tokenizer + model
+tokenizer = AutoTokenizer.from_pretrained(model_id, cache_dir=cache_path)
+
+model = AutoModelForCausalLM.from_pretrained(
+    model_id,
+    torch_dtype=torch.bfloat16,
+    device_map="auto",
+    cache_dir=cache_path,
+)
+
+model.eval()
+print(model)
+
+# -----------------------------
+# 1. Replace ALL linear layers
+# -----------------------------
+def replace_weights_with_random(module):
+    for name, child in module.named_children():
+
+        # If it's a Linear layer, replace weights
+        if isinstance(child, nn.Linear):
+            with torch.no_grad():
+                child.weight.copy_(
+                    torch.randn_like(child.weight)
+                )
+                if child.bias is not None:
+                    child.bias.copy_(
+                        torch.randn_like(child.bias)
+                    )
+
+        else:
+            # recurse into submodules
+            replace_weights_with_random(child)
+
+replace_weights_with_random(model)
+
+print("All linear layer weights replaced with random tensors.")
+
+# -----------------------------
+# 2. Run inference
+# -----------------------------
+prompt = "Explain tensor networks in simple terms."
+
+inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+
+with torch.no_grad():
+    output = model.generate(
+        **inputs,
+        max_new_tokens=500,
+        temperature=1.0,
+        do_sample=True
+    )
+
+print("\n=== OUTPUT ===\n")
+print(tokenizer.decode(output[0], skip_special_tokens=True))
+
+model.save_pretrained("./.cache/qwen_random")
+tokenizer.save_pretrained("./.cache/qwen_random")