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

Add custom kq scaling from Gemma2Attention

Browse source
This commit is contained in:
Andrei Betlen 2024-06-29 10:17:33 -04:00
parent 6f2464e3dd
commit a89427908d
4 changed files with 12 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

3
convert-hf-to-gguf.py
View file

@ -2369,6 +2369,9 @@ class Gemma2Model(Model):
self.gguf_writer.add_final_logit_softcapping(
self.hparams["final_logit_softcapping"]
)
self.gguf_writer.add_query_pre_attn_scalar(
self.hparams["query_pre_attn_scalar"]
)
def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iterable[tuple[str, Tensor]]:
del bid # unusem

1
gguf-py/gguf/constants.py
View file

@ -52,6 +52,7 @@ class Keys:
DECODER_START_TOKEN_ID = "{arch}.decoder_start_token_id"
ATTN_LOGIT_SOFTCAPPING = "{arch}.attn_logit_softcapping"
FINAL_LOGIT_SOFTCAPPING = "{arch}.final_logit_softcapping"
QUERY_PRE_ATTN_SCALAR = "{arch}.query_pre_attn_scalar"
class Attention:
HEAD_COUNT = "{arch}.attention.head_count"

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

@ -522,6 +522,9 @@ class GGUFWriter:
def add_final_logit_softcapping(self, value: float) -> None:
self.add_float32(Keys.LLM.FINAL_LOGIT_SOFTCAPPING.format(arch=self.arch), value)
def add_query_pre_attn_scalar(self, value: float) -> None:
self.add_float32(Keys.LLM.QUERY_PRE_ATTN_SCALAR.format(arch=self.arch), value)
def add_expert_count(self, count: int) -> None:
self.add_uint32(Keys.LLM.EXPERT_COUNT.format(arch=self.arch), count)

6
src/llama.cpp
View file

@ -304,6 +304,7 @@ enum llm_kv {
LLM_KV_DECODER_START_TOKEN_ID,
LLM_KV_ATTN_LOGIT_SOFTCAPPING,
LLM_KV_FINAL_LOGIT_SOFTCAPPING,
LLM_KV_QUERY_PRE_ATTN_SCALAR,
LLM_KV_ATTENTION_HEAD_COUNT,
LLM_KV_ATTENTION_HEAD_COUNT_KV,
@ -396,6 +397,7 @@ static const std::map<llm_kv, const char *> LLM_KV_NAMES = {
{ LLM_KV_DECODER_START_TOKEN_ID, "%s.decoder_start_token_id" },
{ LLM_KV_ATTN_LOGIT_SOFTCAPPING, "%s.attn_logit_softcapping" },
{ LLM_KV_FINAL_LOGIT_SOFTCAPPING, "%s.final_logit_softcapping" },
{ LLM_KV_QUERY_PRE_ATTN_SCALAR, "%s.query_pre_attn_scalar" },
{ LLM_KV_ATTENTION_HEAD_COUNT, "%s.attention.head_count" },
{ LLM_KV_ATTENTION_HEAD_COUNT_KV, "%s.attention.head_count_kv" },
@ -2105,6 +2107,7 @@ struct llama_hparams {
float f_attn_logit_softcapping = 50.0f;
float f_final_logit_softcapping = 30.0f;
float f_query_pre_attn_scalar = 144.0f;
float rope_attn_factor = 1.0f;
float rope_freq_base_train;
@ -4712,6 +4715,7 @@ static void llm_load_hparams(
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
ml.get_key(LLM_KV_ATTN_LOGIT_SOFTCAPPING, hparams.f_attn_logit_softcapping, false);
ml.get_key(LLM_KV_FINAL_LOGIT_SOFTCAPPING, hparams.f_final_logit_softcapping, false);
ml.get_key(LLM_KV_QUERY_PRE_ATTN_SCALAR, hparams.f_query_pre_attn_scalar, false);
hparams.attn_soft_cap = true;
switch (hparams.n_layer) {
@ -10948,7 +10952,7 @@ struct llm_build_context {
ext_factor, attn_factor, beta_fast, beta_slow);
cb(Qcur, "Qcur", il);
Qcur = ggml_scale(ctx0, Qcur, 1.0f / sqrtf(float(n_embd_head_k)));
Qcur = ggml_scale(ctx0, Qcur, 1.0f / sqrtf(hparams.f_query_pre_attn_scalar));
cb(Qcur, "Qcur_scaled", il);
Kcur = ggml_rope_ext(