openai/gpt-oss-20b
Original20.9Bparameters
Fixture26.0Mparameters
Layer stacks1compared
Findings0to review
✓ Same architecture, smaller
Every module class, layer kind and checked property of the original is in the fixture.
Layer stacks
One square per layer, colored by kind: the layer's module classes and type attributes. Hover a square for its index and kind.
model.layers
Original · 24
Fixture · 2
| Kind | Original | Fixture | |
|---|---|---|---|
| sliding_attention | 12 | 1 | ✓ kept |
| full_attention | 12 | 1 | ✓ kept |
Where the parameters are
Share of each model's parameters per component. Vocabulary and patch sizes are kept, so embeddings usually take a larger share of the fixture.
Original
Fixture
| Component | Original | Share | Fixture | Share |
|---|---|---|---|---|
| model.layers | 19.8B | 94.5% | 224.9K | 0.9% |
| lm_head | 579.1M | 2.8% | 12.9M | 49.6% |
| model.embed_tokens | 579.1M | 2.8% | 12.9M | 49.6% |
| model.norm | 2.9K | 0.0% | 64 | 0.0% |
Preserved properties
| Config | Property | Original | Fixture | |
|---|---|---|---|---|
| model | attention style | GQA (64 heads, 8 KV) | GQA (4 heads, 2 KV) | ✓ kept |
| model | MoE routing | top-4 of 32 experts | top-2 of 4 experts | ✓ kept |
| model | RoPE type | yarn | yarn | ✓ kept |
| model | model_type | gpt_oss | gpt_oss | ✓ kept |
| model | vocab_size | 201088 | 201088 | ✓ kept |
| model | tie_word_embeddings | False | False | ✓ kept |
| model | hidden_act | silu | silu | ✓ kept |
| model | max_position_embeddings | 131072 | 131072 | ✓ kept |
Module classes
Instances of each class in the built model (prefix GptOss omitted).
| Class | Original | Fixture |
|---|---|---|
| Linear | 97 | 9 |
| RMSNorm | 49 | 5 |
| Attention | 24 | 2 |
| DecoderLayer | 24 | 2 |
| Experts | 24 | 2 |
| MLP | 24 | 2 |
| TopKRouter | 24 | 2 |
| Embedding | 1 | 1 |
| ForCausalLM | 1 | 1 |
| Model | 1 | 1 |
| RotaryEmbedding | 1 | 1 |
| ModuleList | 1 | 1 |
Config changes
10 fields changed, 29 unchanged.
| Field | Original | Fixture |
|---|---|---|
| head_dim | 64 | 16 |
| hidden_size | 2880 | 64 |
| intermediate_size | 2880 | 128 |
| layer_types | 24 entries: 12× sliding_attention, 12× full_attention | ["sliding_attention", "full_attention"] |
| num_attention_heads | 64 | 4 |
| num_experts_per_tok | 4 | 2 |
| num_hidden_layers | 24 | 2 |
| num_key_value_heads | 8 | 2 |
| num_local_experts | 32 | 4 |
| quantization_config | {"modules_to_not_convert": ["model.layers.*.self_attn", "model.layers.*.mlp.r… | null |