tarekziade HF Staff
Publish MIC usage, slim fixture and reviewed Transformers expectations
d03ef8c verified | { | |
| "mic": "0.3", | |
| "name": "Jamba causal LM", | |
| "usage": "AutoTokenizer + AutoModelForCausalLM, batched, left padding, labels for one training step", | |
| "architecture": "JambaForCausalLM", | |
| "checkpoints": { | |
| "ai21labs/Jamba-tiny-dev": { | |
| "fixture": "huggingface-mic/Jamba-tiny-dev-mic" | |
| } | |
| }, | |
| "fixtures": { | |
| "huggingface-mic/Jamba-tiny-dev-mic": { | |
| "built_from": "ai21labs/Jamba-tiny-dev", | |
| "covers": [ | |
| "ai21labs/Jamba-tiny-dev" | |
| ], | |
| "build": { | |
| "seed": 17 | |
| } | |
| } | |
| }, | |
| "integrations": { | |
| "transformers": { | |
| "entry_point": "integrations/transformers/usage.md", | |
| "format": "runnable-markdown", | |
| "requires": [ | |
| "transformers>=5", | |
| "torch" | |
| ], | |
| "targets": [ | |
| "fixtures", | |
| "checkpoints" | |
| ], | |
| "variants": { | |
| "default": {}, | |
| "eager": { | |
| "attn_implementation": "eager" | |
| }, | |
| "bfloat16": { | |
| "dtype": "bfloat16" | |
| } | |
| }, | |
| "training": true, | |
| "resources": { | |
| "fixture": { | |
| "memory_gb": 2, | |
| "download_mb": 40 | |
| }, | |
| "ai21labs/Jamba-tiny-dev": { | |
| "memory_gb": 8, | |
| "download_mb": 1600 | |
| } | |
| } | |
| } | |
| }, | |
| "outputs": { | |
| "runtime": { | |
| "compare": "exact" | |
| }, | |
| "last_logits": { | |
| "compare": "allclose", | |
| "atol": 0.0001, | |
| "rtol": 0.0001 | |
| }, | |
| "next_token_ids": { | |
| "compare": "exact" | |
| }, | |
| "training.loss": { | |
| "compare": "allclose", | |
| "atol": 0.0001, | |
| "rtol": 0.0001 | |
| }, | |
| "training.gradients": { | |
| "compare": "all_true" | |
| }, | |
| "training.updated": { | |
| "compare": "true" | |
| } | |
| } | |
| } | |