File size: 7,823 Bytes
0c5425e 52570f3 0c5425e e00d697 2cc72c0 e00d697 0c5425e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 | ---
tags:
- mic
- test-artifact
base_model: deepseek-ai/DeepSeek-R1
---
# DeepSeek-R1-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/DeepSeek-R1-mic.html)
## Contract
### DeepSeek-R1-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/DeepSeek-R1-mic.html)
## Contract
### DeepSeek-R1 causal LM
Runnable usage and the framework expectations directory are linked above.
## Slim model
### Slim fixture: DeepSeek-R1-mic
A 16.6M-parameter fixture for the [DeepSeek-R1 MIC](https://huggingface.co/huggingface-mic/DeepSeek-R1-mic), covering [`deepseek-ai/DeepSeek-R1`](https://huggingface.co/deepseek-ai/DeepSeek-R1). It is a test artifact, not a usable model, tagged `test-artifact` to keep it out of candidate-model search.
- **Fixture key `3bd6b377673a`.** It covers every checkpoint whose processor files are byte-identical and whose config matches after shrinking.
- **Same architecture, smaller config:** `DeepseekV3ForCausalLM`, shrunk by the generic rule in [`sweep.py`](https://github.com/huggingface/transformers-mic/blob/main/experiments/sweep.py) (num_hidden_layers 2).
- **Processor files copied byte for byte** from `deepseek-ai/DeepSeek-R1`: `LICENSE`, `configuration_deepseek.py`, `generation_config.json`, `modeling_deepseek.py`, `tokenizer.json`, `tokenizer_config.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`, 0.932), 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 bfloat16**, the dtype `deepseek-ai/DeepSeek-R1` 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 DeepSeek-R1-mic`.
## Slim model
### DeepSeek-R1-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/DeepSeek-R1-mic.html)
## Contract
### DeepSeek-R1-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/DeepSeek-R1-mic.html)
## Contract
### DeepSeek-R1 causal LM
Runnable usage and the framework expectations directory are linked above.
## Slim model
### Slim fixture: DeepSeek-R1-mic
A 16.6M-parameter fixture for the [DeepSeek-R1 MIC](https://huggingface.co/huggingface-mic/DeepSeek-R1-mic), covering [`deepseek-ai/DeepSeek-R1`](https://huggingface.co/deepseek-ai/DeepSeek-R1). It is a test artifact, not a usable model, tagged `test-artifact` to keep it out of candidate-model search.
- **Fixture key `3bd6b377673a`.** It covers every checkpoint whose processor files are byte-identical and whose config matches after shrinking.
- **Same architecture, smaller config:** `DeepseekV3ForCausalLM`, shrunk by the generic rule in [`sweep.py`](https://github.com/huggingface/transformers-mic/blob/main/experiments/sweep.py) (num_hidden_layers 2).
- **Processor files copied byte for byte** from `deepseek-ai/DeepSeek-R1`: `LICENSE`, `configuration_deepseek.py`, `generation_config.json`, `modeling_deepseek.py`, `tokenizer.json`, `tokenizer_config.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`, 0.932), 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 bfloat16**, the dtype `deepseek-ai/DeepSeek-R1` 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 DeepSeek-R1-mic`.
## Slim model
### DeepSeek-R1-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)
## Contract
### DeepSeek-R1 causal LM
Runnable usage and the framework expectations directory are linked above.
## Slim model
### Slim fixture: DeepSeek-R1-mic
A 16.6M-parameter fixture for the [DeepSeek-R1 MIC](https://huggingface.co/huggingface-mic/DeepSeek-R1-mic), covering [`deepseek-ai/DeepSeek-R1`](https://huggingface.co/deepseek-ai/DeepSeek-R1). It is a test artifact, not a usable model, tagged `test-artifact` to keep it out of candidate-model search.
- **Fixture key `3bd6b377673a`.** It covers every checkpoint whose processor files are byte-identical and whose config matches after shrinking.
- **Same architecture, smaller config:** `DeepseekV3ForCausalLM`, shrunk by the generic rule in [`sweep.py`](https://github.com/huggingface/transformers-mic/blob/main/experiments/sweep.py) (num_hidden_layers 2).
- **Processor files copied byte for byte** from `deepseek-ai/DeepSeek-R1`: `LICENSE`, `configuration_deepseek.py`, `generation_config.json`, `modeling_deepseek.py`, `tokenizer.json`, `tokenizer_config.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`, 0.932), 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 bfloat16**, the dtype `deepseek-ai/DeepSeek-R1` 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 DeepSeek-R1-mic`.
|