llama : support glm3 and glm4 (#8031)
* add chatglm3-6b model support huggingface model:
https://hf-mirror.com/THUDM/chatglm3-6b
Signed-off-by: XingXing Qiao <qiaoxx@dingdao.com>
* remove .rotary_pos_emb.inv_freq and unuse code for chatglm3 model
Signed-off-by: XingXing Qiao <qiaoxx@dingdao.com>
* fix lint error
Signed-off-by: XingXing Qiao <qiaoxx@dingdao.com>
* optimize convert-hf-to-gguf.py for chatglm model
Signed-off-by: XingXing Qiao <qiaoxx@dingdao.com>
* support glm-4-9b-chat
Signed-off-by: XingXing Qiao <qiaoxx@dingdao.com>
* fix eos tokens to glm4
* remove unused log
* add preprocess to chatglm3 and chatglm4
* add eos_id_list to llama.cpp
* fix code style
* fix code style
* fix conflicts
* fix conflicts
* Revert "add eos_id_list to llama.cpp"
This reverts commit 3a4d5790bf.
* set <|endoftext|> as eos and <|user|> as eot
* fix chat template bug
* add comment to glm prefix and suffix
* fix conflicts and add rope_ratio & ChatGLMForConditionalGeneration
* fix chat template bug
* fix codestyle
* fix conflicts
* modified the general name of glm model
* fix conflicts
* remove prefix and suffix
* use normal glm4 chattempalte & use LLM_FFN_SWIGLU in phi3
* fix: resolve Flake8 errors in `convert-hf-to-gguf.py`
- Fix E302 by adding two blank lines before top-level function definitions
- Replace print statements to fix NP100
- Fix E303 by ensuring only one blank line between lines of code
* fix rope ratio to solve incorrect answers
* fix by comments
---------
Signed-off-by: XingXing Qiao <qiaoxx@dingdao.com>
Co-authored-by: XingXing Qiao <qiaoxx@dingdao.com>
Co-authored-by: Umpire2018 <138990495+Umpire2018@users.noreply.github.com>
This commit is contained in:
toyer
GitHub
parent
b5040086d4
commit
905942abdb
6 changed files with 455 additions and 25 deletions
247
src/llama.cpp
247
src/llama.cpp
|
|
@ -229,6 +229,7 @@ enum llm_arch {
|
|||
LLM_ARCH_OPENELM,
|
||||
LLM_ARCH_ARCTIC,
|
||||
LLM_ARCH_DEEPSEEK2,
|
||||
LLM_ARCH_CHATGLM,
|
||||
LLM_ARCH_BITNET,
|
||||
LLM_ARCH_T5,
|
||||
LLM_ARCH_JAIS,
|
||||
|
|
@ -272,6 +273,7 @@ static const std::map<llm_arch, const char *> LLM_ARCH_NAMES = {
|
|||
{ LLM_ARCH_OPENELM, "openelm" },
|
||||
{ LLM_ARCH_ARCTIC, "arctic" },
|
||||
{ LLM_ARCH_DEEPSEEK2, "deepseek2" },
|
||||
{ LLM_ARCH_CHATGLM, "chatglm" },
|
||||
{ LLM_ARCH_BITNET, "bitnet" },
|
||||
{ LLM_ARCH_T5, "t5" },
|
||||
{ LLM_ARCH_JAIS, "jais" },
|
||||
|
|
@ -1205,6 +1207,21 @@ static const std::map<llm_arch, std::map<llm_tensor, std::string>> LLM_TENSOR_NA
|
|||
{ LLM_TENSOR_FFN_UP_SHEXP, "blk.%d.ffn_up_shexp" },
|
||||
},
|
||||
},
|
||||
{
|
||||
LLM_ARCH_CHATGLM,
|
||||
{
|
||||
{ LLM_TENSOR_TOKEN_EMBD, "token_embd" },
|
||||
{ LLM_TENSOR_ROPE_FREQS, "rope_freqs" },
|
||||
{ LLM_TENSOR_OUTPUT_NORM, "output_norm" },
|
||||
{ LLM_TENSOR_OUTPUT, "output" },
|
||||
{ LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
|
||||
{ LLM_TENSOR_ATTN_QKV, "blk.%d.attn_qkv" },
|
||||
{ LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
|
||||
{ LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
|
||||
{ LLM_TENSOR_FFN_UP, "blk.%d.ffn_up" },
|
||||
{ LLM_TENSOR_FFN_DOWN, "blk.%d.ffn_down" },
|
||||
},
|
||||
},
|
||||
{
|
||||
LLM_ARCH_BITNET,
|
||||
{
|
||||
|
|
@ -2087,9 +2104,11 @@ enum e_model {
|
|||
MODEL_2_8B,
|
||||
MODEL_3B,
|
||||
MODEL_4B,
|
||||
MODEL_6B,
|
||||
MODEL_6_9B,
|
||||
MODEL_7B,
|
||||
MODEL_8B,
|
||||
MODEL_9B,
|
||||
MODEL_11B,
|
||||
MODEL_12B,
|
||||
MODEL_13B,
|
||||
|
|
@ -2115,7 +2134,6 @@ enum e_model {
|
|||
MODEL_16x12B,
|
||||
MODEL_10B_128x3_66B,
|
||||
MODEL_57B_A14B,
|
||||
MODEL_9B,
|
||||
MODEL_27B,
|
||||
};
|
||||
|
||||
|
|
@ -4490,9 +4508,11 @@ static const char * llama_model_type_name(e_model type) {
|
|||
case MODEL_2_8B: return "2.8B";
|
||||
case MODEL_3B: return "3B";
|
||||
case MODEL_4B: return "4B";
|
||||
case MODEL_6B: return "6B";
|
||||
case MODEL_6_9B: return "6.9B";
|
||||
case MODEL_7B: return "7B";
|
||||
case MODEL_8B: return "8B";
|
||||
case MODEL_9B: return "9B";
|
||||
case MODEL_11B: return "11B";
|
||||
case MODEL_12B: return "12B";
|
||||
case MODEL_13B: return "13B";
|
||||
|
|
@ -4518,7 +4538,6 @@ static const char * llama_model_type_name(e_model type) {
|
|||
case MODEL_16x12B: return "16x12B";
|
||||
case MODEL_10B_128x3_66B: return "10B+128x3.66B";
|
||||
case MODEL_57B_A14B: return "57B.A14B";
|
||||
case MODEL_9B: return "9B";
|
||||
case MODEL_27B: return "27B";
|
||||
default: return "?B";
|
||||
}
|
||||
|
|
@ -5124,6 +5143,15 @@ static void llm_load_hparams(
|
|||
default: model.type = e_model::MODEL_UNKNOWN;
|
||||
}
|
||||
} break;
|
||||
case LLM_ARCH_CHATGLM:
|
||||
{
|
||||
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
|
||||
switch (hparams.n_layer) {
|
||||
case 28: model.type = e_model::MODEL_6B; break;
|
||||
case 40: model.type = e_model::MODEL_9B; break;
|
||||
default: model.type = e_model::MODEL_UNKNOWN;
|
||||
}
|
||||
} break;
|
||||
case LLM_ARCH_BITNET:
|
||||
{
|
||||
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
|
||||
|
|
@ -5256,9 +5284,7 @@ static void llm_load_vocab(
|
|||
if (merges_keyidx == -1) {
|
||||
throw std::runtime_error("cannot find tokenizer merges in model file\n");
|
||||
}
|
||||
|
||||
const int n_merges = gguf_get_arr_n(ctx, merges_keyidx);
|
||||
|
||||
for (int i = 0; i < n_merges; i++) {
|
||||
const std::string word = gguf_get_arr_str(ctx, merges_keyidx, i);
|
||||
GGML_ASSERT(unicode_cpts_from_utf8(word).size() > 0);
|
||||
|
|
@ -5401,6 +5427,10 @@ static void llm_load_vocab(
|
|||
tokenizer_pre == "poro-chat") {
|
||||
vocab.type_pre = LLAMA_VOCAB_PRE_TYPE_PORO;
|
||||
vocab.tokenizer_clean_spaces = false;
|
||||
} else if (
|
||||
tokenizer_pre == "chatglm-bpe") {
|
||||
vocab.type_pre = LLAMA_VOCAB_PRE_TYPE_CHATGLM4;
|
||||
vocab.special_bos_id = -1;
|
||||
} else if (
|
||||
tokenizer_pre == "viking") {
|
||||
vocab.type_pre = LLAMA_VOCAB_PRE_TYPE_VIKING;
|
||||
|
|
@ -5525,7 +5555,6 @@ static void llm_load_vocab(
|
|||
vocab.special_eot_id = 107;
|
||||
}
|
||||
}
|
||||
|
||||
try {
|
||||
vocab.linefeed_id = llama_byte_to_token(vocab, '\n');
|
||||
} catch (const std::exception & e) {
|
||||
|
|
@ -7433,6 +7462,36 @@ static bool llm_load_tensors(
|
|||
layer.ffn_up_b = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_FFN_UP, "bias", i), {n_ff});
|
||||
}
|
||||
} break;
|
||||
case LLM_ARCH_CHATGLM:
|
||||
{
|
||||
model.tok_embd = ml.create_tensor(ctx_input, tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab});
|
||||
|
||||
// output
|
||||
{
|
||||
model.output_norm = ml.create_tensor(ctx_output, tn(LLM_TENSOR_OUTPUT_NORM, "weight"), {n_embd});
|
||||
model.output = ml.create_tensor(ctx_output_split, tn(LLM_TENSOR_OUTPUT, "weight"), {n_embd, n_vocab});
|
||||
}
|
||||
|
||||
for (int i = 0; i < n_layer; ++i) {
|
||||
ggml_context * ctx_layer = ctx_for_layer(i);
|
||||
ggml_context * ctx_split = ctx_for_layer_split(i);
|
||||
|
||||
auto & layer = model.layers[i];
|
||||
|
||||
layer.attn_norm = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_ATTN_NORM, "weight", i), {n_embd});
|
||||
|
||||
layer.wqkv = ml.create_tensor(ctx_split, tn(LLM_TENSOR_ATTN_QKV, "weight", i), {n_embd, n_embd + (hparams.n_embd_head_k << 2)});
|
||||
layer.bqkv = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_ATTN_QKV, "bias", i), {n_embd + (hparams.n_embd_head_k << 2)});
|
||||
|
||||
layer.wo = ml.create_tensor(ctx_split, tn(LLM_TENSOR_ATTN_OUT, "weight", i), {n_embd, n_embd});
|
||||
|
||||
layer.ffn_norm = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_FFN_NORM, "weight", i), {n_embd});
|
||||
|
||||
layer.ffn_up = ml.create_tensor(ctx_split, tn(LLM_TENSOR_FFN_UP, "weight", i), {n_embd, n_ff * 2});
|
||||
|
||||
layer.ffn_down = ml.create_tensor(ctx_split, tn(LLM_TENSOR_FFN_DOWN, "weight", i), {n_ff, n_embd});
|
||||
}
|
||||
} break;
|
||||
default:
|
||||
throw std::runtime_error("unknown architecture");
|
||||
}
|
||||
|
|
@ -7657,6 +7716,7 @@ enum llm_ffn_op_type {
|
|||
LLM_FFN_GELU,
|
||||
LLM_FFN_RELU,
|
||||
LLM_FFN_RELU_SQR,
|
||||
LLM_FFN_SWIGLU,
|
||||
};
|
||||
|
||||
enum llm_ffn_gate_type {
|
||||
|
|
@ -7861,6 +7921,19 @@ static struct ggml_tensor * llm_build_ffn(
|
|||
cur = ggml_sqr(ctx, cur);
|
||||
cb(cur, "ffn_sqr(relu)", il);
|
||||
} break;
|
||||
case LLM_FFN_SWIGLU:
|
||||
{
|
||||
// Project to 4h. If using swiglu double the output width, see https://arxiv.org/pdf/2002.05202.pdf
|
||||
int64_t split_point = cur->ne[0] / 2;
|
||||
struct ggml_tensor * x0 = ggml_cont(ctx, ggml_view_2d(ctx, cur, split_point, cur->ne[1], cur->nb[1], 0));
|
||||
struct ggml_tensor * x1 = ggml_cont(ctx, ggml_view_2d(ctx, cur, split_point, cur->ne[1], cur->nb[1], split_point * ggml_element_size(cur)));
|
||||
|
||||
x0 = ggml_silu(ctx, x0);
|
||||
cb(cur, "ffn_silu", il);
|
||||
|
||||
cur = ggml_mul(ctx, x0, x1);
|
||||
cb(cur, "ffn_mul", il);
|
||||
} break;
|
||||
}
|
||||
|
||||
if (type_gate == LLM_FFN_PAR) {
|
||||
|
|
@ -10709,19 +10782,12 @@ struct llm_build_context {
|
|||
// special-case: the up and gate tensors are merged into a single tensor
|
||||
// TOOD: support into llm_build_ffn
|
||||
{
|
||||
struct ggml_tensor* up = ggml_mul_mat(ctx0, model.layers[il].ffn_up, cur);
|
||||
cb(up, "ffn_up", il);
|
||||
|
||||
auto g = ggml_cont(ctx0, ggml_view_2d(ctx0, up, up->ne[0] / 2, up->ne[1], ggml_row_size(up->type, up->ne[0]), 0));
|
||||
auto y = ggml_cont(ctx0, ggml_view_2d(ctx0, up, up->ne[0] / 2, up->ne[1], ggml_row_size(up->type, up->ne[0]), up->nb[1] / 2));
|
||||
|
||||
y = ggml_mul(ctx0, y, ggml_silu(ctx0, g));
|
||||
cb(y, "ffn_gate", il);
|
||||
|
||||
auto down = ggml_mul_mat(ctx0, model.layers[il].ffn_down, y);
|
||||
cb(down, "ffn_down", il);
|
||||
|
||||
cur = down;
|
||||
cur = llm_build_ffn(ctx0, cur,
|
||||
model.layers[il].ffn_up, NULL, NULL,
|
||||
NULL, NULL, NULL,
|
||||
model.layers[il].ffn_down, NULL, NULL,
|
||||
NULL,
|
||||
LLM_FFN_SWIGLU, LLM_FFN_SEQ, cb, il);
|
||||
cb(cur, "ffn_out", il);
|
||||
}
|
||||
|
||||
|
|
@ -13413,6 +13479,120 @@ struct llm_build_context {
|
|||
|
||||
return gf;
|
||||
}
|
||||
|
||||
struct ggml_cgraph * build_chatglm() {
|
||||
struct ggml_cgraph * gf = ggml_new_graph_custom(ctx0, LLAMA_MAX_NODES, false);
|
||||
|
||||
const int64_t n_embd_head = hparams.n_embd_head_v;
|
||||
const int64_t n_embd_gqa = hparams.n_embd_v_gqa();
|
||||
GGML_ASSERT(n_embd_head == hparams.n_embd_head_k);
|
||||
|
||||
struct ggml_tensor * cur;
|
||||
struct ggml_tensor * inpL;
|
||||
|
||||
inpL = llm_build_inp_embd(ctx0, lctx, hparams, batch, model.tok_embd, cb);
|
||||
|
||||
// inp_pos - contains the positions
|
||||
struct ggml_tensor * inp_pos = build_inp_pos();
|
||||
|
||||
// KQ_mask (mask for 1 head, it will be broadcasted to all heads)
|
||||
struct ggml_tensor * KQ_mask = build_inp_KQ_mask();
|
||||
|
||||
for (int il = 0; il < n_layer; ++il) {
|
||||
struct ggml_tensor * inpSA = inpL;
|
||||
|
||||
cur = llm_build_norm(ctx0, inpL, hparams,
|
||||
model.layers[il].attn_norm,
|
||||
NULL,
|
||||
LLM_NORM_RMS, cb, il);
|
||||
cb(cur, "attn_norm", il);
|
||||
|
||||
// self-attention
|
||||
{
|
||||
struct ggml_tensor * Qcur = nullptr;
|
||||
struct ggml_tensor * Kcur = nullptr;
|
||||
struct ggml_tensor * Vcur = nullptr;
|
||||
|
||||
cur = ggml_mul_mat(ctx0, model.layers[il].wqkv, cur);
|
||||
cb(cur, "wqkv", il);
|
||||
|
||||
cur = ggml_add(ctx0, cur, model.layers[il].bqkv);
|
||||
cb(cur, "bqkv", il);
|
||||
|
||||
Qcur = ggml_cont(ctx0, ggml_view_2d(ctx0, cur, n_embd, n_tokens, cur->nb[1], 0*sizeof(float)*(n_embd)));
|
||||
Kcur = ggml_cont(ctx0, ggml_view_2d(ctx0, cur, n_embd_gqa, n_tokens, cur->nb[1], 1*sizeof(float)*(n_embd)));
|
||||
Vcur = ggml_cont(ctx0, ggml_view_2d(ctx0, cur, n_embd_gqa, n_tokens, cur->nb[1], 1*sizeof(float)*(n_embd + n_embd_gqa)));
|
||||
|
||||
cb(Qcur, "Qcur", il);
|
||||
cb(Kcur, "Kcur", il);
|
||||
cb(Vcur, "Vcur", il);
|
||||
//printf("freq_base: %f freq_scale: %f ext_factor: %f attn_factor: %f\n", freq_base, freq_scale, ext_factor, attn_factor);
|
||||
Qcur = ggml_rope_ext(
|
||||
ctx0, ggml_reshape_3d(ctx0, Qcur, n_embd_head, n_head, n_tokens), inp_pos, nullptr,
|
||||
n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,
|
||||
ext_factor, attn_factor, beta_fast, beta_slow
|
||||
);
|
||||
cb(Qcur, "Qcur_rope", il);
|
||||
|
||||
Kcur = ggml_rope_ext(
|
||||
ctx0, ggml_reshape_3d(ctx0, Kcur, n_embd_head, n_head_kv, n_tokens), inp_pos, nullptr,
|
||||
n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,
|
||||
ext_factor, attn_factor, beta_fast, beta_slow
|
||||
);
|
||||
cb(Kcur, "Kcur_rope", il);
|
||||
|
||||
cur = llm_build_kv(ctx0, model, hparams, cparams, kv_self, gf,
|
||||
model.layers[il].wo, NULL,
|
||||
Kcur, Vcur, Qcur, KQ_mask, n_tokens, kv_head, n_kv, 1.0f/sqrtf(float(n_embd_head)), cb, il);
|
||||
|
||||
}
|
||||
|
||||
if (il == n_layer - 1) {
|
||||
// skip computing output for unused tokens
|
||||
struct ggml_tensor * inp_out_ids = build_inp_out_ids();
|
||||
cur = ggml_get_rows(ctx0, cur, inp_out_ids);
|
||||
inpSA = ggml_get_rows(ctx0, inpSA, inp_out_ids);
|
||||
}
|
||||
|
||||
// Add the input
|
||||
struct ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA);
|
||||
cb(ffn_inp, "ffn_inp", il);
|
||||
|
||||
// FF
|
||||
{
|
||||
cur = llm_build_norm(ctx0, ffn_inp, hparams,
|
||||
model.layers[il].ffn_norm,
|
||||
NULL,
|
||||
LLM_NORM_RMS, cb, il);
|
||||
cb(cur, "ffn_norm", il);
|
||||
|
||||
cur = llm_build_ffn(ctx0, cur,
|
||||
model.layers[il].ffn_up, NULL, NULL,
|
||||
NULL, NULL, NULL,
|
||||
model.layers[il].ffn_down, NULL, NULL,
|
||||
NULL,
|
||||
LLM_FFN_SWIGLU, LLM_FFN_SEQ, cb, il);
|
||||
cb(cur, "ffn_out", il);
|
||||
|
||||
}
|
||||
|
||||
inpL = ggml_add(ctx0, cur, ffn_inp);
|
||||
cb(inpL, "l_out", il);
|
||||
}
|
||||
|
||||
cur = llm_build_norm(ctx0, inpL, hparams,
|
||||
model.output_norm,
|
||||
NULL,
|
||||
LLM_NORM_RMS, cb, -1);
|
||||
cb(cur, "result_norm", -1);
|
||||
|
||||
cur = ggml_mul_mat(ctx0, model.output, cur);
|
||||
cb(cur, "result_output", -1);
|
||||
|
||||
ggml_build_forward_expand(gf, cur);
|
||||
|
||||
return gf;
|
||||
}
|
||||
};
|
||||
|
||||
static struct ggml_cgraph * llama_build_graph_defrag(llama_context & lctx, const std::vector<uint32_t> & ids) {
|
||||
|
|
@ -13644,6 +13824,10 @@ static struct ggml_cgraph * llama_build_graph(
|
|||
{
|
||||
result = llm.build_deepseek2();
|
||||
} break;
|
||||
case LLM_ARCH_CHATGLM:
|
||||
{
|
||||
result = llm.build_chatglm();
|
||||
} break;
|
||||
case LLM_ARCH_BITNET:
|
||||
{
|
||||
result = llm.build_bitnet();
|
||||
|
|
@ -15259,6 +15443,11 @@ struct llm_tokenizer_bpe {
|
|||
" ?[^(\\s|.,!?…。，、।۔،)]+",
|
||||
};
|
||||
break;
|
||||
case LLAMA_VOCAB_PRE_TYPE_CHATGLM4:
|
||||
regex_exprs = {
|
||||
"(?:'[sS]|'[tT]|'[rR][eE]|'[vV][eE]|'[mM]|'[lL][lL]|'[dD])|[^\\r\\n\\p{L}\\p{N}]?\\p{L}+|\\p{N}{1,3}| ?[^\\s\\p{L}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
|
||||
};
|
||||
break;
|
||||
case LLAMA_VOCAB_PRE_TYPE_VIKING:
|
||||
regex_exprs = {
|
||||
"\\p{N}",
|
||||
|
|
@ -16160,7 +16349,6 @@ static std::vector<llama_vocab::id> llama_tokenize_internal(const llama_vocab &
|
|||
if (add_special) {
|
||||
tokenizer.append_bos(output);
|
||||
}
|
||||
|
||||
for (const auto & fragment : fragment_buffer) {
|
||||
if (fragment.type == FRAGMENT_BUFFER_VARIANT_TYPE_RAW_TEXT) {
|
||||
auto raw_text = fragment.raw_text.substr(fragment.offset, fragment.length);
|
||||
|
|
@ -19151,6 +19339,7 @@ enum llama_rope_type llama_rope_type(const struct llama_model * model) {
|
|||
case LLM_ARCH_OLMO:
|
||||
case LLM_ARCH_ARCTIC:
|
||||
case LLM_ARCH_DEEPSEEK2:
|
||||
case LLM_ARCH_CHATGLM:
|
||||
return LLAMA_ROPE_TYPE_NORM;
|
||||
|
||||
// the pairs of head values are offset by n_rot/2
|
||||
|
|
@ -20883,7 +21072,6 @@ int32_t llama_tokenize(
|
|||
bool add_special,
|
||||
bool parse_special) {
|
||||
auto res = llama_tokenize_internal(model->vocab, std::string(text, text_len), add_special, parse_special);
|
||||
|
||||
if (n_tokens_max < (int) res.size()) {
|
||||
// LLAMA_LOG_ERROR("%s: too many tokens\n", __func__);
|
||||
return -((int) res.size());
|
||||
|
|
@ -21302,6 +21490,25 @@ static int32_t llama_chat_apply_template_internal(
|
|||
if (add_ass) {
|
||||
ss << "<|start_header_id|>assistant<|end_header_id|>\n\n";
|
||||
}
|
||||
} else if (tmpl == "chatglm3" || tmpl_contains("[gMASK]sop")) {
|
||||
// chatglm3-6b
|
||||
ss << "[gMASK]" << "sop";
|
||||
for (auto message : chat) {
|
||||
std::string role(message->role);
|
||||
ss << "<|" << role << "|>" << "\n " << message->content;
|
||||
}
|
||||
if (add_ass) {
|
||||
ss << "<|assistant|>";
|
||||
}
|
||||
} else if (tmpl == "chaglm4" || tmpl_contains("[gMASK]<sop>")) {
|
||||
ss << "[gMASK]" << "<sop>";
|
||||
for (auto message : chat) {
|
||||
std::string role(message->role);
|
||||
ss << "<|" << role << "|>" << "\n" << message->content;
|
||||
}
|
||||
if (add_ass) {
|
||||
ss << "<|assistant|>";
|
||||
}
|
||||
} else if (tmpl == "minicpm" || tmpl_contains(u8"<用户>")) {
|
||||
// MiniCPM-3B-OpenHermes-2.5-v2-GGUF
|
||||
for (auto message : chat) {
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue