vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

update: support Qwen2-57B-A14B (#7835)

Browse source 
* update: convert-hf-to-gguf.py to support Qwen2-57B-A14B

* fix: QWEN2MOE support for expert_feed_forward_length

previously, expert ff was taken from n_ff (intermediate size) but it is now properly taken from LLM_KV_EXPERT_FEED_FORWARD_LENGTH

n_ff_exp and n_ff_shared_exp are now properly calculated

* update: convert-hf-to-gguf.py cleanup for Qwen2MoeForCausalLM

* fix: QWEN2MOE support for expert_feed_forward_length

previously, expert ff was taken from n_ff (intermediate size) but it is now properly taken from LLM_KV_EXPERT_FEED_FORWARD_LENGTH

n_ff_exp and n_ff_shexp are now properly calculated
This commit is contained in:
Ștefan-Gabriel Muscalu 2024-06-17 22:08:46 +03:00 • committed by GitHub
parent 5b6da18750
commit a94e6ff877
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
4 changed files with 57 additions and 34 deletions
Download patch file Download diff file Expand all files Collapse all files

31
gguf-py/gguf/constants.py
View file

@ -33,21 +33,22 @@ class Keys:
FILE_TYPE = "general.file_type"
class LLM:
VOCAB_SIZE = "{arch}.vocab_size"
CONTEXT_LENGTH = "{arch}.context_length"
EMBEDDING_LENGTH = "{arch}.embedding_length"
BLOCK_COUNT = "{arch}.block_count"
LEADING_DENSE_BLOCK_COUNT = "{arch}.leading_dense_block_count"
FEED_FORWARD_LENGTH = "{arch}.feed_forward_length"
EXPERT_FEED_FORWARD_LENGTH = "{arch}.expert_feed_forward_length"
USE_PARALLEL_RESIDUAL = "{arch}.use_parallel_residual"
TENSOR_DATA_LAYOUT = "{arch}.tensor_data_layout"
EXPERT_COUNT = "{arch}.expert_count"
EXPERT_USED_COUNT = "{arch}.expert_used_count"
EXPERT_SHARED_COUNT = "{arch}.expert_shared_count"
EXPERT_WEIGHTS_SCALE = "{arch}.expert_weights_scale"
POOLING_TYPE = "{arch}.pooling_type"
LOGIT_SCALE = "{arch}.logit_scale"
VOCAB_SIZE = "{arch}.vocab_size"
CONTEXT_LENGTH = "{arch}.context_length"
EMBEDDING_LENGTH = "{arch}.embedding_length"
BLOCK_COUNT = "{arch}.block_count"
LEADING_DENSE_BLOCK_COUNT = "{arch}.leading_dense_block_count"
FEED_FORWARD_LENGTH = "{arch}.feed_forward_length"
EXPERT_FEED_FORWARD_LENGTH = "{arch}.expert_feed_forward_length"
EXPERT_SHARED_FEED_FORWARD_LENGTH = "{arch}.expert_shared_feed_forward_length"
USE_PARALLEL_RESIDUAL = "{arch}.use_parallel_residual"
TENSOR_DATA_LAYOUT = "{arch}.tensor_data_layout"
EXPERT_COUNT = "{arch}.expert_count"
EXPERT_USED_COUNT = "{arch}.expert_used_count"
EXPERT_SHARED_COUNT = "{arch}.expert_shared_count"
EXPERT_WEIGHTS_SCALE = "{arch}.expert_weights_scale"
POOLING_TYPE = "{arch}.pooling_type"
LOGIT_SCALE = "{arch}.logit_scale"
class Attention:
HEAD_COUNT = "{arch}.attention.head_count"

3
gguf-py/gguf/gguf_writer.py
View file

@ -394,6 +394,9 @@ class GGUFWriter:
def add_expert_feed_forward_length(self, length: int) -> None:
self.add_uint32(Keys.LLM.EXPERT_FEED_FORWARD_LENGTH.format(arch=self.arch), length)
def add_expert_shared_feed_forward_length(self, length: int) -> None:
self.add_uint32(Keys.LLM.EXPERT_SHARED_FEED_FORWARD_LENGTH.format(arch=self.arch), length)
def add_parallel_residual(self, use: bool) -> None:
self.add_bool(Keys.LLM.USE_PARALLEL_RESIDUAL.format(arch=self.arch), use)