tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-08-14 20:29:41 -04:00
Code Issues Packages Projects Releases Wiki Activity

llama : Support intern-s1 (#14875)

Browse Source 
* support internvl

* support interns1

* resolve comments

* put interns1 in tensor mapping

* resolve comment

* move tokenizer changes to sub class
This commit is contained in:
RunningLeon
2025-08-08 00:20:40 +08:00
committed by GitHub
parent 7ad67ba9fe
commit 99acbc9921
2 changed files with 122 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
gguf-py/gguf/tensor_mapping.py
View File

@@ -1110,11 +1110,13 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_EMBD_CLS: (
"vision_tower.vision_model.embeddings.class_embedding",
"model.vision_tower.embeddings.cls_token", # Intern-S1
"vision_model.class_embedding", # llama 4
),
MODEL_TENSOR.V_ENC_EMBD_PATCH: (
"vision_tower.vision_model.embeddings.patch_embedding",
"model.vision_tower.embeddings.patch_embeddings.projection", # Intern-S1
"vpm.embeddings.patch_embedding",
"model.vision_model.embeddings.patch_embedding", # SmolVLM
"vision_tower.patch_conv", # pixtral
@@ -1124,6 +1126,7 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_EMBD_POS: (
"vision_tower.vision_model.embeddings.position_embedding",
"model.vision_tower.embeddings.position_embeddings", # Intern-S1
"vpm.embeddings.position_embedding",
"model.vision_model.embeddings.position_embedding", # SmolVLM
"vision_model.positional_embedding_vlm", # llama 4
@@ -1131,6 +1134,7 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_ATTN_Q: (
"vision_tower.vision_model.encoder.layers.{bid}.self_attn.q_proj",
"model.vision_tower.encoder.layer.{bid}.attention.q_proj", # Intern-S1
"vpm.encoder.layers.{bid}.self_attn.q_proj",
"model.vision_model.encoder.layers.{bid}.self_attn.q_proj", # SmolVLM
"vision_model.model.layers.{bid}.self_attn.q_proj", # llama4
@@ -1140,10 +1144,12 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_ATTN_Q_NORM: (
"vision_tower.vision_model.encoder.layers.{bid}.attn.q_norm", # InternVL
"model.vision_tower.encoder.layer.{bid}.attention.q_norm", # Intern-S1
),
MODEL_TENSOR.V_ENC_ATTN_K: (
"vision_tower.vision_model.encoder.layers.{bid}.self_attn.k_proj",
"model.vision_tower.encoder.layer.{bid}.attention.k_proj", # Intern-S1
"vpm.encoder.layers.{bid}.self_attn.k_proj",
"model.vision_model.encoder.layers.{bid}.self_attn.k_proj", # SmolVLM
"vision_model.model.layers.{bid}.self_attn.k_proj", # llama4
@@ -1153,10 +1159,12 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_ATTN_K_NORM: (
"vision_tower.vision_model.encoder.layers.{bid}.attn.k_norm", # InternVL
"model.vision_tower.encoder.layer.{bid}.attention.k_norm", # Intern-S1
),
MODEL_TENSOR.V_ENC_ATTN_V: (
"vision_tower.vision_model.encoder.layers.{bid}.self_attn.v_proj",
"model.vision_tower.encoder.layer.{bid}.attention.v_proj", # Intern-S1
"vpm.encoder.layers.{bid}.self_attn.v_proj",
"model.vision_model.encoder.layers.{bid}.self_attn.v_proj", # SmolVLM
"vision_model.model.layers.{bid}.self_attn.v_proj", # llama4
@@ -1167,6 +1175,7 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_INPUT_NORM: (
"vision_tower.vision_model.encoder.layers.{bid}.layer_norm1",
"vision_tower.vision_model.encoder.layers.{bid}.norm1", # InternVL
"model.vision_tower.encoder.layer.{bid}.layernorm_before", # Intern-S1
"vpm.encoder.layers.{bid}.layer_norm1",
"model.vision_model.encoder.layers.{bid}.layer_norm1", # SmolVLM
"vision_tower.transformer.layers.{bid}.attention_norm", # pixtral
@@ -1177,6 +1186,7 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_ATTN_O: (
"vision_tower.vision_model.encoder.layers.{bid}.self_attn.out_proj",
"vision_tower.vision_model.encoder.layers.{bid}.attn.proj", # InternVL
"model.vision_tower.encoder.layer.{bid}.attention.projection_layer", # Intern-S1
"vpm.encoder.layers.{bid}.self_attn.out_proj",
"model.vision_model.encoder.layers.{bid}.self_attn.out_proj", # SmolVLM
"vision_model.model.layers.{bid}.self_attn.o_proj", # llama4
@@ -1187,6 +1197,7 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_POST_ATTN_NORM: (
"vision_tower.vision_model.encoder.layers.{bid}.layer_norm2",
"vision_tower.vision_model.encoder.layers.{bid}.norm2", # InternVL
"model.vision_tower.encoder.layer.{bid}.layernorm_after", # Intern-S1
"vpm.encoder.layers.{bid}.layer_norm2",
"model.vision_model.encoder.layers.{bid}.layer_norm2", # SmolVLM
"vision_model.model.layers.{bid}.post_attention_layernorm", # llama4
@@ -1196,6 +1207,7 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_FFN_UP: (
"vision_tower.vision_model.encoder.layers.{bid}.mlp.fc1",
"model.vision_tower.encoder.layer.{bid}.mlp.fc1", # Intern-S1
"vpm.encoder.layers.{bid}.mlp.fc1",
"model.vision_model.encoder.layers.{bid}.mlp.fc1", # SmolVLM, gemma3
"vision_tower.transformer.layers.{bid}.feed_forward.up_proj", # pixtral
@@ -1211,6 +1223,7 @@ class TensorNameMap:
MODEL_TENSOR.V_ENC_FFN_DOWN: (
"vision_tower.vision_model.encoder.layers.{bid}.mlp.fc2",
"model.vision_tower.encoder.layer.{bid}.mlp.fc2", # Intern-S1
"vpm.encoder.layers.{bid}.mlp.fc2",
"model.vision_model.encoder.layers.{bid}.mlp.fc2", # SmolVLM, gemma3
"vision_tower.transformer.layers.{bid}.feed_forward.down_proj", # pixtral
@@ -1221,10 +1234,12 @@ class TensorNameMap:
MODEL_TENSOR.V_LAYER_SCALE_1: (
"vision_tower.vision_model.encoder.layers.{bid}.ls1", # InternVL
"model.vision_tower.encoder.layer.{bid}.lambda_1", # Intern-S1
),
MODEL_TENSOR.V_LAYER_SCALE_2: (
"vision_tower.vision_model.encoder.layers.{bid}.ls2", # InternVL
"model.vision_tower.encoder.layer.{bid}.lambda_2", # Intern-S1
),
MODEL_TENSOR.V_PRE_NORM: (