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

llama : add PLaMo model (#3557)

Browse source 
* add plamo mock

* add tensor loading

* plamo convert

* update norm

* able to compile

* fix norm_rms_eps hparam

* runnable

* use inp_pos

* seems ok

* update kqv code

* remove develop code

* update README

* shuffle attn_q.weight and attn_output.weight for broadcasting

* remove plamo_llm_build_kqv and use llm_build_kqv

* fix style

* update

* llama : remove obsolete KQ_scale

* plamo : fix tensor names for correct GPU offload

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Shintarou Okada 2023-12-24 22:35:49 +09:00 • committed by GitHub
parent 5bf3953d7e
commit 753be377b6
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
5 changed files with 307 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

17
gguf-py/gguf/constants.py
View file

@ -96,6 +96,7 @@ class MODEL_ARCH(IntEnum):
STABLELM = auto()
QWEN = auto()
PHI2 = auto()
PLAMO = auto()
class MODEL_TENSOR(IntEnum):
@ -142,6 +143,7 @@ MODEL_ARCH_NAMES: dict[MODEL_ARCH, str] = {
MODEL_ARCH.STABLELM: "stablelm",
MODEL_ARCH.QWEN: "qwen",
MODEL_ARCH.PHI2: "phi2",
MODEL_ARCH.PLAMO: "plamo",
}
TENSOR_NAMES: dict[MODEL_TENSOR, str] = {
@ -349,6 +351,21 @@ MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = {
MODEL_TENSOR.FFN_DOWN,
MODEL_TENSOR.FFN_UP,
],
MODEL_ARCH.PLAMO: [
MODEL_TENSOR.TOKEN_EMBD,
MODEL_TENSOR.OUTPUT_NORM,
MODEL_TENSOR.OUTPUT,
MODEL_TENSOR.ROPE_FREQS,
MODEL_TENSOR.ATTN_NORM,
MODEL_TENSOR.ATTN_Q,
MODEL_TENSOR.ATTN_K,
MODEL_TENSOR.ATTN_V,
MODEL_TENSOR.ATTN_OUT,
MODEL_TENSOR.ATTN_ROT_EMBD,
MODEL_TENSOR.FFN_GATE,
MODEL_TENSOR.FFN_DOWN,
MODEL_TENSOR.FFN_UP,
],
MODEL_ARCH.GPT2: [
# TODO
],

37
gguf-py/gguf/tensor_mapping.py
View file

@ -79,6 +79,7 @@ class TensorNameMap:
"language_model.encoder.layers.{bid}.input_layernorm", # persimmon
"model.layers.{bid}.ln1", # yi
"transformer.h.{bid}.ln", # phi2
"model.layers.layers.{bid}.norm", # plamo
),
# Attention norm 2
@ -99,26 +100,29 @@ class TensorNameMap:
# Attention query
MODEL_TENSOR.ATTN_Q: (
"model.layers.{bid}.self_attn.q_proj", # llama-hf
"layers.{bid}.attention.wq", # llama-pth
"encoder.layer.{bid}.attention.self.query", # bert
"transformer.h.{bid}.attn.q_proj", # gpt-j
"model.layers.{bid}.self_attn.q_proj", # llama-hf
"layers.{bid}.attention.wq", # llama-pth
"encoder.layer.{bid}.attention.self.query", # bert
"transformer.h.{bid}.attn.q_proj", # gpt-j
"model.layers.layers.{bid}.self_attn.q_proj", # plamo
),
# Attention key
MODEL_TENSOR.ATTN_K: (
"model.layers.{bid}.self_attn.k_proj", # llama-hf
"layers.{bid}.attention.wk", # llama-pth
"encoder.layer.{bid}.attention.self.key", # bert
"transformer.h.{bid}.attn.k_proj", # gpt-j
"model.layers.{bid}.self_attn.k_proj", # llama-hf
"layers.{bid}.attention.wk", # llama-pth
"encoder.layer.{bid}.attention.self.key", # bert
"transformer.h.{bid}.attn.k_proj", # gpt-j
"model.layers.layers.{bid}.self_attn.k_proj", # plamo
),
# Attention value
MODEL_TENSOR.ATTN_V: (
"model.layers.{bid}.self_attn.v_proj", # llama-hf
"layers.{bid}.attention.wv", # llama-pth
"encoder.layer.{bid}.attention.self.value", # bert
"transformer.h.{bid}.attn.v_proj", # gpt-j
"model.layers.{bid}.self_attn.v_proj", # llama-hf
"layers.{bid}.attention.wv", # llama-pth
"encoder.layer.{bid}.attention.self.value", # bert
"transformer.h.{bid}.attn.v_proj", # gpt-j
"model.layers.layers.{bid}.self_attn.v_proj", # plamo
),
# Attention output
@ -134,12 +138,14 @@ class TensorNameMap:
"transformer.h.{bid}.attn.out_proj", # gpt-j
"language_model.encoder.layers.{bid}.self_attention.dense", # persimmon
"transformer.h.{bid}.mixer.out_proj", # phi2
"model.layers.layers.{bid}.self_attn.o_proj", # plamo
),
# Rotary embeddings
MODEL_TENSOR.ATTN_ROT_EMBD: (
"model.layers.{bid}.self_attn.rotary_emb.inv_freq", # llama-hf
"layers.{bid}.attention.inner_attention.rope.freqs", # llama-pth
"model.layers.{bid}.self_attn.rotary_emb.inv_freq", # llama-hf
"layers.{bid}.attention.inner_attention.rope.freqs", # llama-pth
"model.layers.layers.{bid}.self_attn.rotary_emb.inv_freq", # plamo
),
# Feed-forward norm
@ -174,6 +180,7 @@ class TensorNameMap:
"language_model.encoder.layers.{bid}.mlp.dense_h_to_4h", # persimmon
"transformer.h.{bid}.mlp.w1", # qwen
"transformer.h.{bid}.mlp.fc1", # phi2
"model.layers.layers.{bid}.mlp.up_proj", # plamo
),
MODEL_TENSOR.FFN_UP_EXP: (
@ -186,6 +193,7 @@ class TensorNameMap:
"model.layers.{bid}.mlp.gate_proj", # llama-hf refact
"layers.{bid}.feed_forward.w1", # llama-pth
"transformer.h.{bid}.mlp.w2", # qwen
"model.layers.layers.{bid}.mlp.gate_proj", # plamo
),
MODEL_TENSOR.FFN_GATE_EXP: (
@ -206,6 +214,7 @@ class TensorNameMap:
"transformer.h.{bid}.mlp.fc_out", # gpt-j
"language_model.encoder.layers.{bid}.mlp.dense_4h_to_h", # persimmon
"transformer.h.{bid}.mlp.fc2", # phi2
"model.layers.layers.{bid}.mlp.down_proj", # plamo
),
MODEL_TENSOR.FFN_DOWN_EXP: (