vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-11 09:34:37 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'upstream' into concedo_experimental

Browse source 
# Conflicts:
#	README.md
#	docs/build.md
#	docs/development/HOWTO-add-model.md
#	tests/test-backend-ops.cpp
#	tests/test-chat-template.cpp
This commit is contained in:
Concedo 2025-01-10 17:57:38 +08:00
commit b154bd3671
50 changed files with 189817 additions and 187510 deletions
Download patch file Download diff file Expand all files Collapse all files

19
src/llama-model.cpp
View file

@ -80,6 +80,7 @@ const char * llm_type_name(llm_type type) {
case MODEL_8x7B: return "8x7B";
case MODEL_8x22B: return "8x22B";
case MODEL_16x12B: return "16x12B";
case MODEL_16x3_8B: return "16x3.8B";
case MODEL_10B_128x3_66B: return "10B+128x3.66B";
case MODEL_57B_A14B: return "57B.A14B";
case MODEL_27B: return "27B";
@ -668,6 +669,15 @@ void llm_load_hparams(llama_model_loader & ml, llama_model & model) {
throw std::runtime_error("invalid value for sliding_window");
}
} break;
case LLM_ARCH_PHIMOE:
{
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
switch (hparams.n_layer) {
case 32: model.type = e_model::MODEL_16x3_8B; break;
default: model.type = e_model::MODEL_UNKNOWN;
}
} break;
case LLM_ARCH_PLAMO:
{
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
@ -1051,12 +1061,15 @@ void llm_load_hparams(llama_model_loader & ml, llama_model & model) {
}
} break;
case LLM_ARCH_RWKV6:
case LLM_ARCH_RWKV6QWEN2:
{
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_EPS, hparams.f_norm_eps);
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_EPS, hparams.f_norm_eps, false);
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps, false);
ml.get_key(LLM_KV_WKV_HEAD_SIZE, hparams.wkv_head_size);
ml.get_key(LLM_KV_TIME_MIX_EXTRA_DIM, hparams.time_mix_extra_dim);
ml.get_key(LLM_KV_TIME_DECAY_EXTRA_DIM, hparams.time_decay_extra_dim);
ml.get_key(LLM_KV_RESCALE_EVERY_N_LAYERS, hparams.rescale_every_n_layers, false);
ml.get_key(LLM_KV_TOKEN_SHIFT_COUNT, hparams.token_shift_count, false);
switch (hparams.n_layer) {
case 24: model.type = e_model::MODEL_1_6B; break;
@ -1067,6 +1080,7 @@ void llm_load_hparams(llama_model_loader & ml, llama_model & model) {
default: model.type = e_model::MODEL_UNKNOWN;
} break;
case 61: model.type = e_model::MODEL_14B; break;
case 64: model.type = e_model::MODEL_32B; break;
default: model.type = e_model::MODEL_UNKNOWN;
}
} break;
@ -2074,6 +2088,7 @@ enum llama_rope_type llama_rope_type(const struct llama_model * model) {
case LLM_ARCH_T5ENCODER:
case LLM_ARCH_JAIS:
case LLM_ARCH_RWKV6:
case LLM_ARCH_RWKV6QWEN2:
case LLM_ARCH_WAVTOKENIZER_DEC:
return LLAMA_ROPE_TYPE_NONE;
@ -2114,6 +2129,7 @@ enum llama_rope_type llama_rope_type(const struct llama_model * model) {
case LLM_ARCH_OLMOE:
case LLM_ARCH_PHI2:
case LLM_ARCH_PHI3:
case LLM_ARCH_PHIMOE:
case LLM_ARCH_GEMMA:
case LLM_ARCH_GEMMA2:
case LLM_ARCH_STARCODER2:
@ -2217,6 +2233,7 @@ bool llama_model_is_recurrent(const struct llama_model * model) {
switch (model->arch) {
case LLM_ARCH_MAMBA: return true;
case LLM_ARCH_RWKV6: return true;
case LLM_ARCH_RWKV6QWEN2: return true;
default: return false;
}
}