codegen2-16B_P-mic / README.md
tarekziade's picture
tarekziade HF Staff
Publish MIC usage, slim fixture and reviewed Transformers expectations
1fea908 verified
|
Raw History Blame Contribute Delete
2.43 kB
---
tags:
- mic
- test-artifact
base_model: Salesforce/codegen2-16B_P
---
# codegen2-16B_P-mic
Slim model weights, executable usage and framework expectations in one repository.
- [Contract manifest](integration.json)
- [Transformers usage](integrations/transformers/usage.md)
- [Transformers expectations](integrations/transformers/expectations)
- [Fixture provenance](build.json)
- [How the fixture was shrunk](https://huggingface-mic-mic-coverage.static.hf.space/shrink/codegen2-16B_P-mic.html)
## Contract
### CodeGen2 text generation
Runnable usage and the framework expectations directory are linked above.
## Slim model
### Slim fixture: codegen2-16B_P-mic
A 6.7M-parameter fixture for the [CodeGen2 MIC](https://huggingface.co/huggingface-mic/codegen2-16B_P-mic), covering [`Salesforce/codegen2-16B_P`](https://huggingface.co/Salesforce/codegen2-16B_P). It is a test artifact, not a usable model, tagged `test-artifact` to keep it out of candidate-model search.
- **Fixture key `3423601d143b`.** It covers every checkpoint whose processor files are byte-identical and whose config matches after shrinking.
- **Same architecture, smaller config:** `CodeGenForCausalLM`, shrunk by the generic rule in [`sweep.py`](https://github.com/huggingface/transformers-mic/blob/main/experiments/sweep.py) ().
- **Processor files copied byte for byte** from `Salesforce/codegen2-16B_P`: `added_tokens.json`, `configuration_codegen.py`, `merges.txt`, `modeling_codegen.py`, `special_tokens_map.json`, `tokenizer.json`, `tokenizer_config.json`, `vocab.json`.
- **Random weights (seed 17).** They passed the input-sensitivity gate: every perturbation of the contract's inputs moved the logits by at least 0.001 (weakest: `input_ids:roll_batch`, 1.04), so no training was needed.
- **Same tensor names as the real checkpoint.** The real checkpoint has every parameter, and so does the fixture.
- **Saved in float32**, the dtype `Salesforce/codegen2-16B_P` loads in by default.
- **Finite in float32, bfloat16, and float16** (forward, loss, gradients).
Built by [`build_fixtures.py`](https://github.com/huggingface/transformers-mic/blob/main/example/build_fixtures.py); [`build.json`](build.json) records the pinned sources, gate results, and file hashes. This repository includes the slim weights, usage and framework expectations; rebuild the other files with `python example/build_fixtures.py --fixture codegen2-16B_P-mic`.