mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-08-15 20:53:00 -04:00
model: support GLM 4.5 family of models (#14939)
* model: Add GLM 4.5 (#14921) Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * Merge in PR suggestions Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * model: Add GLM 4.5 family of models (#14921) 1. Updated tensor_mapping.py with NextN tensor mappings - Added proper tensor mappings for all NextN/MTP tensors in /Users/samm/git/llama.cpp/gguf-py/gguf/tensor_mapping.py - Added mappings for: eh_proj, embed_tokens, enorm, hnorm, shared_head.head, shared_head.norm 2. Added num_nextn_predict_layers configuration - Added LLM_KV_NUM_NEXTN_PREDICT_LAYERS constant to llama-arch.h and llama-arch.cpp - Added num_nextn_predict_layers field to llama_hparams struct - Updated GLM4_MOE parameter loading in llama-model.cpp to read this parameter - Modified tensor loading logic to conditionally load NextN tensors based on num_nextn_predict_layers - Added GGUF writer support in gguf_writer.py with add_num_nextn_predict_layers() method - Updated conversion script to extract and write this parameter from HuggingFace config 3. Added FIM tokens for GLM4_MOE - Added GLM-4.5's FIM tokens to llama-vocab.cpp: - <|code_prefix|> for FIM_PRE - <|code_suffix|> for FIM_SUF - <|code_middle|> for FIM_MID 4. Removed manual NextN tensor handling - Removed the special-case handling in convert_hf_to_gguf.py that manually mapped NextN tensors - NextN tensors are now handled automatically through the proper tensor mapping system * glm 4.5 update tensors names * model: glm 4.5 apply suggestions from code review Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * Update src/llama-model.cpp Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * model: glm 4.5 apply suggestions from code review Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * model: glm 4.5 apply suggestions from code review * Apply suggestions from code review * patch broken chat template * typings fix * add TENSOR_SKIP flag Co-authored-by: Diego Devesa <slarengh@gmail.com> * Update src/llama-model-loader.h Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> --------- Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> Co-authored-by: Diego Devesa <slarengh@gmail.com>
This commit is contained in:
Sam
GitHub
@@ -105,6 +105,7 @@ class Keys:
|
||||
EXPERT_WEIGHTS_NORM = "{arch}.expert_weights_norm"
|
||||
EXPERT_GATING_FUNC = "{arch}.expert_gating_func"
|
||||
MOE_EVERY_N_LAYERS = "{arch}.moe_every_n_layers"
|
||||
NEXTN_PREDICT_LAYERS = "{arch}.nextn_predict_layers"
|
||||
POOLING_TYPE = "{arch}.pooling_type"
|
||||
LOGIT_SCALE = "{arch}.logit_scale"
|
||||
DECODER_START_TOKEN_ID = "{arch}.decoder_start_token_id"
|
||||
@@ -357,6 +358,7 @@ class MODEL_ARCH(IntEnum):
|
||||
DEEPSEEK2 = auto()
|
||||
CHATGLM = auto()
|
||||
GLM4 = auto()
|
||||
GLM4_MOE = auto()
|
||||
BITNET = auto()
|
||||
T5 = auto()
|
||||
T5ENCODER = auto()
|
||||
@@ -614,6 +616,13 @@ class MODEL_TENSOR(IntEnum):
|
||||
A_MMPROJ_FC = auto()
|
||||
A_MM_NORM_PRE = auto()
|
||||
A_MM_NORM_MID = auto()
|
||||
# nextn/mtp
|
||||
NEXTN_EH_PROJ = auto()
|
||||
NEXTN_EMBED_TOKENS = auto()
|
||||
NEXTN_ENORM = auto()
|
||||
NEXTN_HNORM = auto()
|
||||
NEXTN_SHARED_HEAD_HEAD = auto()
|
||||
NEXTN_SHARED_HEAD_NORM = auto()
|
||||
|
||||
|
||||
MODEL_ARCH_NAMES: dict[MODEL_ARCH, str] = {
|
||||
@@ -678,6 +687,7 @@ MODEL_ARCH_NAMES: dict[MODEL_ARCH, str] = {
|
||||
MODEL_ARCH.DEEPSEEK2: "deepseek2",
|
||||
MODEL_ARCH.CHATGLM: "chatglm",
|
||||
MODEL_ARCH.GLM4: "glm4",
|
||||
MODEL_ARCH.GLM4_MOE: "glm4moe",
|
||||
MODEL_ARCH.BITNET: "bitnet",
|
||||
MODEL_ARCH.T5: "t5",
|
||||
MODEL_ARCH.T5ENCODER: "t5encoder",
|
||||
@@ -936,6 +946,13 @@ TENSOR_NAMES: dict[MODEL_TENSOR, str] = {
|
||||
MODEL_TENSOR.A_MMPROJ_FC: "mm.a.fc",
|
||||
MODEL_TENSOR.A_MM_NORM_PRE: "mm.a.norm_pre",
|
||||
MODEL_TENSOR.A_MM_NORM_MID: "mm.a.norm_mid",
|
||||
# NextN/MTP
|
||||
MODEL_TENSOR.NEXTN_EH_PROJ: "blk.{bid}.nextn.eh_proj",
|
||||
MODEL_TENSOR.NEXTN_EMBED_TOKENS: "blk.{bid}.nextn.embed_tokens",
|
||||
MODEL_TENSOR.NEXTN_ENORM: "blk.{bid}.nextn.enorm",
|
||||
MODEL_TENSOR.NEXTN_HNORM: "blk.{bid}.nextn.hnorm",
|
||||
MODEL_TENSOR.NEXTN_SHARED_HEAD_HEAD: "blk.{bid}.nextn.shared_head_head",
|
||||
MODEL_TENSOR.NEXTN_SHARED_HEAD_NORM: "blk.{bid}.nextn.shared_head_norm",
|
||||
}
|
||||
|
||||
MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = {
|
||||
@@ -2124,6 +2141,37 @@ MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = {
|
||||
MODEL_TENSOR.ATTN_POST_NORM,
|
||||
MODEL_TENSOR.FFN_POST_NORM,
|
||||
],
|
||||
MODEL_ARCH.GLM4_MOE: [
|
||||
MODEL_TENSOR.TOKEN_EMBD,
|
||||
MODEL_TENSOR.OUTPUT_NORM,
|
||||
MODEL_TENSOR.OUTPUT,
|
||||
MODEL_TENSOR.ATTN_NORM,
|
||||
MODEL_TENSOR.ATTN_POST_NORM,
|
||||
MODEL_TENSOR.ATTN_Q,
|
||||
MODEL_TENSOR.ATTN_K,
|
||||
MODEL_TENSOR.ATTN_V,
|
||||
MODEL_TENSOR.ATTN_OUT,
|
||||
MODEL_TENSOR.ATTN_Q_NORM,
|
||||
MODEL_TENSOR.ATTN_K_NORM,
|
||||
MODEL_TENSOR.FFN_GATE,
|
||||
MODEL_TENSOR.FFN_DOWN,
|
||||
MODEL_TENSOR.FFN_UP,
|
||||
MODEL_TENSOR.FFN_GATE_INP,
|
||||
MODEL_TENSOR.FFN_GATE_EXP,
|
||||
MODEL_TENSOR.FFN_DOWN_EXP,
|
||||
MODEL_TENSOR.FFN_UP_EXP,
|
||||
MODEL_TENSOR.FFN_GATE_SHEXP,
|
||||
MODEL_TENSOR.FFN_DOWN_SHEXP,
|
||||
MODEL_TENSOR.FFN_UP_SHEXP,
|
||||
MODEL_TENSOR.FFN_EXP_PROBS_B,
|
||||
# NextN/MTP tensors - preserved but unused
|
||||
MODEL_TENSOR.NEXTN_EH_PROJ,
|
||||
MODEL_TENSOR.NEXTN_EMBED_TOKENS,
|
||||
MODEL_TENSOR.NEXTN_ENORM,
|
||||
MODEL_TENSOR.NEXTN_HNORM,
|
||||
MODEL_TENSOR.NEXTN_SHARED_HEAD_HEAD,
|
||||
MODEL_TENSOR.NEXTN_SHARED_HEAD_NORM,
|
||||
],
|
||||
MODEL_ARCH.BITNET: [
|
||||
MODEL_TENSOR.ATTN_Q,
|
||||
MODEL_TENSOR.ATTN_K,
|
||||
|
||||
Reference in New Issue
Block a user