Add Granite 4.1 20B support (GraniteSWA adapter) #283
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,215 @@ | ||
| # Copyright 2025 The Torch-Spyre Authors. | ||
| # | ||
| # Licensed under the Apache License, Version 2.0 (the "License"); | ||
| # you may not use this file except in compliance with the License. | ||
| # You may obtain a copy of the License at | ||
| # | ||
| # http://www.apache.org/licenses/LICENSE-2.0 | ||
| # | ||
| # Unless required by applicable law or agreed to in writing, software | ||
| # distributed under the License is distributed on an "AS IS" BASIS, | ||
| # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
| # See the License for the specific language governing permissions and | ||
| # limitations under the License. | ||
|
|
||
| """ | ||
| HuggingFace Transformers adapter for Granite 4.1 SWA models on Spyre. | ||
|
|
||
| GraniteSWAForCausalLM is identical to Granite 3.x except that layers alternate | ||
| between full attention and sliding-window attention (``layer_type`` attribute on | ||
| each ``GraniteSWADecoderLayer``). Sliding-window layers receive a local attention | ||
| mask that restricts each token to attend only within the ``sliding_window`` window; | ||
| full-attention layers receive the normal causal mask. | ||
|
|
||
| Usage:: | ||
|
|
||
| from hf_adapters import AutoSpyreModelForCausalLM | ||
| from transformers import AutoTokenizer | ||
|
|
||
| model = AutoSpyreModelForCausalLM.from_pretrained( | ||
| "/tmp/models/granite-4.1-20b") | ||
| tokenizer = AutoTokenizer.from_pretrained("/tmp/models/granite-4.1-20b") | ||
| outputs = model.generate(tokenizer, ["Hello!"], max_new_tokens=32) | ||
| """ | ||
|
|
||
| import torch | ||
| import torch.nn.functional as F | ||
|
|
||
| from hf_adapters.hf_common import ( | ||
| apply_rope_matmul, | ||
| get_backbone, | ||
| kv_cache_update, | ||
| pad_lm_head, | ||
| patch_rmsnorm, | ||
| prepare_rope_and_heads, | ||
| ) | ||
|
|
||
|
|
||
| def _make_compiled_block(layer, sliding_window: int): | ||
| """Compiled block for GraniteSWA. | ||
|
|
||
| Full-attention layers receive the normal causal ``attn_mask``. | ||
| Sliding-window layers receive a local mask built from the same buffer but | ||
| restricted to a ``sliding_window``-wide band; positions outside the window | ||
| are masked to ``-inf``. | ||
| """ | ||
| attn = layer.self_attn | ||
| mlp = layer.mlp | ||
| input_ln = layer.input_layernorm | ||
| post_attn_ln = layer.post_attention_layernorm | ||
| res_mult = layer.residual_multiplier | ||
| v_head_dim = getattr(attn, "v_head_dim", attn.head_dim) | ||
| is_sliding = ( | ||
| getattr(layer, "layer_type", "full_attention") == "sliding_attention" | ||
| ) | ||
|
|
||
| def block_forward( | ||
| hidden_states, | ||
| selected_freqs, | ||
| attn_mask, | ||
| key_cache, | ||
| value_cache, | ||
| is_filling, | ||
| token_index, | ||
| cache_position, | ||
| ): | ||
| residual = hidden_states | ||
| h = input_ln(hidden_states) | ||
|
|
||
| bsz, seq_len, _ = h.shape | ||
| q = attn.q_proj(h).view(bsz, seq_len, -1, attn.head_dim).transpose(1, 2) | ||
| k = attn.k_proj(h).view(bsz, seq_len, -1, attn.head_dim).transpose(1, 2) | ||
| v = attn.v_proj(h).view(bsz, seq_len, -1, v_head_dim).transpose(1, 2) | ||
|
|
||
| q = apply_rope_matmul(q, selected_freqs) | ||
| k = apply_rope_matmul(k, selected_freqs) | ||
|
|
||
| key_cache, value_cache = kv_cache_update( | ||
| k, | ||
| v, | ||
| key_cache, | ||
| value_cache, | ||
| is_filling, | ||
| token_index, | ||
| cache_position, | ||
| ) | ||
|
|
||
| # For sliding-window layers, apply a local band mask on top of the | ||
| # causal mask: positions further than sliding_window steps back are | ||
| # set to -inf so the model never attends outside the window. | ||
| if is_sliding: | ||
| cache_len = key_cache.shape[2] | ||
| q_len = q.shape[2] | ||
| # Anchor query positions to the real token positions so that the | ||
| # window is correct during both prefill and single-token decode. | ||
| # During prefill token_index=0 and q_len>1; during decode | ||
| # token_index is the current position and q_len=1. | ||
| q_pos = (token_index + torch.arange(q_len, device=q.device)).unsqueeze(1) # [q, 1] | ||
| k_pos = torch.arange(cache_len, device=q.device).unsqueeze(0) # [1, k] | ||
| window_mask = (q_pos - k_pos) >= sliding_window # [q, k] | ||
| swa_mask = attn_mask.clone() | ||
| swa_mask = swa_mask.masked_fill( | ||
| window_mask.unsqueeze(0).unsqueeze(0), float("-inf") | ||
| ) | ||
| effective_mask = swa_mask | ||
| else: | ||
| effective_mask = attn_mask | ||
|
|
||
| attn_out = F.scaled_dot_product_attention( | ||
| q, | ||
| key_cache, | ||
| value_cache, | ||
| attn_mask=effective_mask, | ||
| dropout_p=0.0, | ||
| scale=attn.scaling, | ||
| enable_gqa=True, | ||
| ) | ||
| attn_out = attn_out.transpose(1, 2).reshape(bsz, seq_len, -1) | ||
| attn_out = attn.o_proj(attn_out) | ||
|
|
||
| h = residual + attn_out * res_mult | ||
|
|
||
| residual = h | ||
| h = post_attn_ln(h) | ||
| h = mlp(h) | ||
| h = residual + h * res_mult | ||
|
|
||
| return h, key_cache, value_cache | ||
|
|
||
| return torch.compile(block_forward, dynamic=False) | ||
|
|
||
|
|
||
| def _run_backbone_forward( | ||
|
Contributor
There was a problem hiding this comment. Maybe better to do more reuse of the existing adapter, like in #281 |
||
| model, | ||
| input_ids, | ||
| position_ids, | ||
| attn_mask, | ||
| key_caches, | ||
| value_caches, | ||
| is_filling, | ||
| token_index, | ||
| cache_position, | ||
| ): | ||
| """GraniteSWA backbone: embedding * multiplier, blocks, norm.""" | ||
| backbone = get_backbone(model) | ||
| h = backbone.embed_tokens(input_ids) | ||
| h = h * backbone.embedding_multiplier | ||
|
|
||
| selected_freqs = model._spyre_rope(h, position_ids) | ||
|
|
||
| for i, compiled_block in enumerate(model._spyre_compiled_blocks): | ||
| h, key_caches[i], value_caches[i] = compiled_block( | ||
| h, | ||
| selected_freqs, | ||
| attn_mask, | ||
| key_caches[i], | ||
| value_caches[i], | ||
| is_filling, | ||
| token_index, | ||
| cache_position, | ||
| ) | ||
|
|
||
| h = backbone.norm(h) | ||
| return h | ||
|
|
||
|
|
||
| def _run_forward( | ||
| model, | ||
| input_ids, | ||
| position_ids, | ||
| attn_mask, | ||
| key_caches, | ||
| value_caches, | ||
| is_filling, | ||
| token_index, | ||
| cache_position, | ||
| ): | ||
| """GraniteSWA causal-LM forward: backbone + lm_head / logits_scaling.""" | ||
| h = _run_backbone_forward( | ||
| model, | ||
| input_ids, | ||
| position_ids, | ||
| attn_mask, | ||
| key_caches, | ||
| value_caches, | ||
| is_filling, | ||
| token_index, | ||
| cache_position, | ||
| ) | ||
| logits = model.lm_head(h) | ||
| logits = logits / model.config.logits_scaling | ||
| return logits | ||
|
|
||
|
|
||
| def prepare_for_spyre(model): | ||
| """Apply Spyre adaptations to a GraniteSWA model in-place.""" | ||
| from transformers.models.granite_swa.modeling_granite_swa import GraniteSWARMSNorm | ||
|
|
||
| sliding_window = model.config.sliding_window | ||
| prepare_rope_and_heads(model) | ||
| patch_rmsnorm(GraniteSWARMSNorm) | ||
| pad_lm_head(model) | ||
| model._spyre_compiled_blocks = [ | ||
| _make_compiled_block(layer, sliding_window) | ||
| for layer in get_backbone(model).layers | ||
| ] | ||
Oops, something went wrong.
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The import currently fails, so to merge this change we would need to wait for a transformers release that includes this config