tarekziade HF Staff
Publish MIC usage, slim fixture and reviewed Transformers expectations
0c5425e verified | { | |
| "recorded_with": { | |
| "contract": "huggingface-mic/DeepSeek-R1-mic", | |
| "contract_name": "DeepSeek-R1", | |
| "contract_revision": "ecb2ceeafd962e7f11e6c5ce69874528e26b8619", | |
| "framework": "transformers", | |
| "variant": "eager", | |
| "target": "DeepSeek-R1-mic", | |
| "kind": "fixture", | |
| "revision": "ecb2ceeafd962e7f11e6c5ce69874528e26b8619", | |
| "environment": { | |
| "python": "3.11.18", | |
| "transformers": "5.22.0.dev0", | |
| "torch": "2.14.0+cu130", | |
| "torchvision": "0.29.2", | |
| "torchaudio": "2.11.0+cu130", | |
| "tokenizers": "0.23.3", | |
| "numpy": "2.4.6", | |
| "pillow": "13.0.0", | |
| "accelerate": "1.16.0.dev0", | |
| "safetensors": "0.8.0", | |
| "huggingface-hub": "2.3.0", | |
| "sentencepiece": "0.2.2", | |
| "protobuf": "7.36.2", | |
| "librosa": "0.11.0", | |
| "soundfile": "0.14.0", | |
| "scipy": "1.17.1" | |
| }, | |
| "seconds": 7.92, | |
| "environment_differs": { | |
| "torchvision": "0.29.2 (pinned 0.29.1)", | |
| "numpy": "2.4.6 (pinned 2.3.5)", | |
| "pillow": "13.0.0 (pinned 12.3.0)", | |
| "accelerate": "1.16.0.dev0 (pinned 1.15.0)" | |
| }, | |
| "platform": "cuda" | |
| }, | |
| "outputs": { | |
| "runtime": { | |
| "device": "cuda", | |
| "dtype": "bfloat16", | |
| "attn_implementation": "eager" | |
| }, | |
| "next_token_ids": [ | |
| 119704, | |
| 5475 | |
| ], | |
| "probe_logits": [ | |
| [ | |
| -0.1787109375, | |
| 0.126953125, | |
| 0.0390625, | |
| 0.036865234375, | |
| 0.01611328125 | |
| ], | |
| [ | |
| -0.2119140625, | |
| 0.1201171875, | |
| -0.01324462890625, | |
| 0.07080078125, | |
| -0.0238037109375 | |
| ] | |
| ], | |
| "training": { | |
| "loss": 11.722162246704102, | |
| "gradients": { | |
| "model.layers.0.self_attn.q_a_proj.weight": true, | |
| "model.layers.0.mlp.gate_proj.weight": true | |
| }, | |
| "updated": true | |
| } | |
| } | |
| } | |