llama : add StarCoder2 support (#5795)

* Add support for starcoder2

* handle rope type

* skip rope freq and rotary embeddings from being serialized

* resolve comments

* Update llama.cpp

* remove redundant changes

* handle `rope-theta`

* llama : change starcoder2 rope type

* address comment

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Sourab Mangrulkar 2024-03-02 01:00:46 +05:30 committed by GitHub
parent 38d16b1426
commit c29af7e225
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
4 changed files with 229 additions and 1 deletions

View file

@ -96,9 +96,11 @@ class Model:
if (n_head_kv := self.hparams.get("num_key_value_heads")) is not None:
self.gguf_writer.add_head_count_kv(n_head_kv)
if (rope_theta := self.hparams.get("rope_theta")) is not None:
self.gguf_writer.add_rope_freq_base(rope_theta)
if (f_rms_eps := self.hparams.get("rms_norm_eps")) is not None:
self.gguf_writer.add_layer_norm_rms_eps(f_rms_eps)
if (f_norm_eps := self.find_hparam(["layer_norm_eps", "layer_norm_epsilon"], optional=True)) is not None:
if (f_norm_eps := self.find_hparam(["layer_norm_eps", "layer_norm_epsilon", "norm_epsilon"], optional=True)) is not None:
self.gguf_writer.add_layer_norm_eps(f_norm_eps)
if (n_experts := self.hparams.get("num_local_experts")) is not None:
self.gguf_writer.add_expert_count(n_experts)
@ -220,6 +222,8 @@ class Model:
return NomicBertModel
if model_architecture == "GemmaForCausalLM":
return GemmaModel
if model_architecture == "Starcoder2ForCausalLM":
return Model
return Model
def _is_model_safetensors(self) -> bool:
@ -281,6 +285,8 @@ class Model:
return gguf.MODEL_ARCH.NOMIC_BERT
if arch == "GemmaForCausalLM":
return gguf.MODEL_ARCH.GEMMA
if arch == "Starcoder2ForCausalLM":
return gguf.MODEL_ARCH.STARCODER2
raise NotImplementedError(f'Architecture "{arch}" not supported!')