happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

fixed logical errors in convert_hf_to_gguf.py pointed out by CISC

This commit is contained in:
Yee Man Chan 2026-02-03 08:14:21 +08:00
parent 11282a0f60
commit 4bb4286f7d
1 changed files with 26 additions and 36 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

62
convert_hf_to_gguf.py
View File

@ -5088,7 +5088,7 @@ class KimiLinearModel(TextModel):
# KDA & MLA params
# Get ssm_d_conv from linear_attn_config.short_conv_kernel_size or ssm_d_conv
linear_attn_config = self.find_hparam(["linear_attn_config"], optional=False)
linear_attn_config = self.hparams["linear_attn_config"]
# n_head == 0 for KDA layers, n_head > 0 for MLA layers
# full_attention_layers list will be used to distingush layer type
_num_kv_heads = list()
@ -5108,57 +5108,47 @@ class KimiLinearModel(TextModel):
# MLA params - use add_* methods that handle arch substitution
# Support both HuggingFace naming (q_lora_rank, kv_lora_rank) and internal naming (n_lora_q, n_lora_kv)
if (q_lora_rank := self.find_hparam(["q_lora_rank", "n_lora_q"], optional=False)) is not None:
if (q_lora_rank := self.find_hparam(["q_lora_rank", "n_lora_q"], optional=True)) is not None:
self.gguf_writer.add_q_lora_rank(q_lora_rank)
# To enable MLA KV cache, MLA needs to be converted into MQA with larger heads, then decompresses to MHA
if (kv_lora_rank := self.find_hparam(["kv_lora_rank", "n_lora_kv"], optional=False)) is not None:
self.gguf_writer.add_kv_lora_rank(kv_lora_rank)
# MLA head dimensions
# Support HuggingFace naming: qk_nope_head_dim, qk_rope_head_dim, v_head_dim
qk_nope_head_dim = self.find_hparam(["qk_nope_head_dim"], optional=False)
qk_rope_head_dim = self.find_hparam(["qk_rope_head_dim"], optional=False)
v_head_dim = self.find_hparam(["v_head_dim"], optional=False)
kv_lora_rank = self.find_hparam(["kv_lora_rank"], optional=False)
# To enable MLA KV cache, MLA needs to be converted into MQA with larger heads, then decompresses to MHA
self.gguf_writer.add_key_length(kv_lora_rank + qk_rope_head_dim)
self.gguf_writer.add_value_length(kv_lora_rank)
# Calculate n_embd_head_k_mla = qk_nope_head_dim + qk_rope_head_dim
if (n_embd_head_k_mla := self.find_hparam(["n_embd_head_k_mla"], optional=True)) is not None:
self.gguf_writer.add_key_length_mla(n_embd_head_k_mla)
elif qk_nope_head_dim is not None and qk_rope_head_dim is not None:
n_embd_head_k_mla = qk_nope_head_dim + qk_rope_head_dim
self.gguf_writer.add_key_length_mla(n_embd_head_k_mla)
# n_embd_head_v_mla = v_head_dim
if (n_embd_head_v_mla := self.find_hparam(["n_embd_head_v_mla"], optional=True)) is not None:
self.gguf_writer.add_value_length_mla(n_embd_head_v_mla)
elif v_head_dim is not None:
self.gguf_writer.add_value_length_mla(v_head_dim)
qk_nope_head_dim = self.hparams.get("qk_nope_head_dim")
# Rotation - use qk_rope_head_dim for Kimi
if (rope_dim := self.find_hparam(["qk_rope_head_dim", "n_rot"], optional=True)) is not None:
self.gguf_writer.add_rope_dimension_count(rope_dim)
if (qk_rope_head_dim := self.find_hparam(["qk_rope_head_dim", "n_rot"], optional=False)) is not None:
self.gguf_writer.add_rope_dimension_count(qk_rope_head_dim)
self.gguf_writer.add_key_length(kv_lora_rank + qk_rope_head_dim)
else:
# Default to head_dim
head_dim = self.hparams["hidden_size"] // self.hparams["num_attention_heads"]
self.gguf_writer.add_rope_dimension_count(head_dim)
self.gguf_writer.add_key_length(kv_lora_rank + head_dim)
v_head_dim = self.hparams.get("v_head_dim")
# Calculate n_embd_head_k_mla = qk_nope_head_dim + qk_rope_head_dim
if (n_embd_head_k_mla := self.find_hparam(["n_embd_head_k_mla"], optional=True)) is not None:
self.gguf_writer.add_key_length_mla(n_embd_head_k_mla)
elif qk_nope_head_dim is not None:
n_embd_head_k_mla = qk_nope_head_dim + qk_rope_head_dim
self.gguf_writer.add_key_length_mla(n_embd_head_k_mla)
# n_embd_head_v_mla = v_head_dim
if (n_embd_head_v_mla := self.hparams.get("n_embd_head_v_mla")) is not None:
self.gguf_writer.add_value_length_mla(n_embd_head_v_mla)
elif v_head_dim is not None:
self.gguf_writer.add_value_length_mla(v_head_dim)
# moe_intermediate_size (1024 for Kimi)
if (moe_intermediate_size := self.find_hparam(["moe_intermediate_size"], optional=False)) is not None:
self.gguf_writer.add_expert_feed_forward_length(moe_intermediate_size)
self.gguf_writer.add_expert_feed_forward_length(self.hparams["moe_intermediate_size"])
# num_shared_experts (1 for Kimi)
if (num_shared_experts := self.find_hparam(["num_shared_experts"], optional=False)) is not None:
self.gguf_writer.add_expert_shared_count(num_shared_experts)
self.gguf_writer.add_expert_shared_count(self.hparams["num_shared_experts"])
# first_k_dense_replace (1 for Kimi - first layer uses dense MLP)
if (first_k_dense_replace := self.find_hparam(["first_k_dense_replace"])) is not None:
self.gguf_writer.add_leading_dense_block_count(first_k_dense_replace)
self.gguf_writer.add_leading_dense_block_count(self.hparams["first_k_dense_replace"])
# Routed scaling factor (expert_weights_scale = 2.446 for Kimi)
if (routed_scaling_factor := self.find_hparam(["routed_scaling_factor"], optional=False)) is not None:
self.gguf_writer.add_expert_weights_scale(routed_scaling_factor)
self.gguf_writer.add_expert_weights_scale(self.hparams["routed_scaling_factor"])
def prepare_tensors(self):
super().prepare_tensors()