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

Fix issues during StableLM models conversion

Browse source
This commit is contained in:
Anas Ahouzi 2024-02-24 10:37:44 -08:00
parent 9e359a4f47
commit 8a529990fa
1 changed files with 3 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

6
convert-hf-to-gguf.py
View file

@ -192,7 +192,7 @@ class Model:
return RefactModel
if model_architecture == "PersimmonForCausalLM":
return PersimmonModel
if model_architecture in ("StableLMEpochForCausalLM", "LlavaStableLMEpochForCausalLM"):
if model_architecture in ("StableLmForCausalLM", "LlavaStableLMEpochForCausalLM"):
return StableLMModel
if model_architecture == "QWenLMHeadModel":
return QwenModel
@ -253,7 +253,7 @@ class Model:
return gguf.MODEL_ARCH.REFACT
if arch == "PersimmonForCausalLM":
return gguf.MODEL_ARCH.PERSIMMON
if arch in ("StableLMEpochForCausalLM", "LlavaStableLMEpochForCausalLM"):
if arch in ("StableLmForCausalLM", "LlavaStableLMEpochForCausalLM"):
return gguf.MODEL_ARCH.STABLELM
if arch == "QWenLMHeadModel":
return gguf.MODEL_ARCH.QWEN
@ -1074,7 +1074,7 @@ class StableLMModel(Model):
self.gguf_writer.add_embedding_length(hparams["hidden_size"])
self.gguf_writer.add_block_count(block_count)
self.gguf_writer.add_feed_forward_length(hparams["intermediate_size"])
self.gguf_writer.add_rope_dimension_count(int(hparams["rope_pct"] * (hparams["hidden_size"] // hparams["num_attention_heads"])))
self.gguf_writer.add_rope_dimension_count(int(hparams["partial_rotary_factor"] * (hparams["hidden_size"] // hparams["num_attention_heads"])))
self.gguf_writer.add_head_count(hparams["num_attention_heads"])
self.gguf_writer.add_parallel_residual(hparams["use_parallel_residual"] if "use_parallel_residual" in hparams else True)
self.gguf_writer.add_layer_norm_eps(1e-5)