MCPcopy Create free account

hub / github.com/Facico/Chinese-Vicuna / functions

Functions461 in github.com/Facico/Chinese-Vicuna

⨍ Functions 461◇ Types & classes 80↳ Endpoints 1
Functionggml_relu
tools/Vicuna.cpp/ggml.c:3508
Functionggml_relu_inplace
tools/Vicuna.cpp/ggml.c:3514
Functionggml_reshape
tools/Vicuna.cpp/ggml.c:3767
Functionggml_reshape_2d
tools/Vicuna.cpp/ggml.c:3792
Functionggml_rms_norm_inplace
tools/Vicuna.cpp/ggml.c:3650
Functionggml_scale_inplace
tools/Vicuna.cpp/ggml.c:3717
Functionggml_set_i32_1d
tools/Vicuna.cpp/ggml.c:2866
Functionggml_set_param
tools/Vicuna.cpp/ggml.c:4214
Functionggml_set_scratch
tools/Vicuna.cpp/ggml.c:2492
Functionggml_sgn_inplace
tools/Vicuna.cpp/ggml.c:3412
Functionggml_silu_inplace
tools/Vicuna.cpp/ggml.c:3582
Functionggml_sqr
tools/Vicuna.cpp/ggml.c:3231
Functionggml_sqr_inplace
tools/Vicuna.cpp/ggml.c:3237
Functionggml_sqrt
tools/Vicuna.cpp/ggml.c:3265
Functionggml_sqrt_inplace
tools/Vicuna.cpp/ggml.c:3271
Functionggml_step_inplace
tools/Vicuna.cpp/ggml.c:3480
Functionggml_sub
tools/Vicuna.cpp/ggml.c:3109
Functionggml_sub_inplace
tools/Vicuna.cpp/ggml.c:3116
Functionggml_time_ms
tools/Vicuna.cpp/ggml.c:295
Functionggml_up64
tools/Vicuna.cpp/ggml.c:2356
Functionggml_vec_silu_f16
tools/Vicuna.cpp/ggml.c:1948
Functionggml_view_2d
tools/Vicuna.cpp/ggml.c:3868
Functiongpt_random_prompt
tools/Vicuna.cpp/utils.cpp:107
Functiongpt_tokenize
tools/Vicuna.cpp/utils.cpp:225
Functiongpt_vocab_init
tools/Vicuna.cpp/utils.cpp:318
Functioninteraction
( input, history, temperature=0.1, top_p=0.75, top_k=40, num_beams=4, max_new_toke
interaction.py:111
Methodkernel_call
()
tools/quant/custom_autotune.py:63
Functionload_quant
(model, checkpoint, wbits, groupsize=-1, fused_mlp=True, eval=True, warmup_autotune=True)
tools/quant_llama.py:246
Functionmain
usage: ./llama-quantize models/llama/ggml-model.bin models/llama/ggml-model-quant.bin type
tools/Vicuna.cpp/quantize.cpp:291
Functionmain
tools/Vicuna.cpp/chat.cpp:799
Functionmake_fused_mlp
Replace all LlamaMLP modules with QuantLlamaMLP modules, which fuses many of the operations.
tools/quant/fused_mlp.py:235
Functionmake_quant_attn
Replace all LlamaAttention modules with QuantLlamaAttention modules, fusing the q, k, v projections.
tools/quant/fused_attn.py:86
Functionmake_quant_linear
(module, names, bits, groupsize, name='')
tools/quant/quant_linear.py:397
Functionmatmul248_kernel_config_pruner
The main purpose of this function is to shrink BLOCK_SIZE_* when the corresponding dimension is smaller.
tools/quant/custom_autotune.py:167
Functionmatmul_248_kernel
Compute the matrix multiplication C = A x B. A is of shape (M, K) float16 B is of shape (K//8, N) int32 C is of s
tools/quant/quant_linear.py:72
Functionnoop
(*args, **kwargs)
tools/quant_llama.py:250
Functionnoop
(*args, **kwargs)
tools/quant_generate.py:30
Methodon_log
(self, args, state, control, logs, **kwargs)
finetune_chat.py:233
Functionpermute
(w)
tools/merge_lora.py:102
Methodpostprocess
(self, text, render=True)
prompt.py:78
Methodpreprocess_gen
(self, data_point)
prompt.py:24
Methodpreprocess_train
(self, data_point)
prompt.py:39
Methodpreprocess_train
(self, data_point)
prompt.py:152
Functionpthread_create
tools/Vicuna.cpp/ggml.c:53
Functionpthread_join
tools/Vicuna.cpp/ggml.c:64
Functionsave
( inputs, history, temperature=0.1, top_p=0.75, top_k=40, num_beams=4, max_new_tok
chat.py:122
Functionset_console_logger
(name)
utils.py:64
Functionset_file_logger
(name, dir, use_console=False)
utils.py:73
Functionset_seed
(seed)
tools/datautils.py:5
Functionsigint_handler
tools/Vicuna.cpp/chat.cpp:767
Functionskip
(*args, **kwargs)
tools/quant_llama.py:21
Functiontmp
(_, inp, out)
tools/quant_llama.py:105
Functionto_json
(data, path)
utils.py:102
Functionto_jsonl
(data, path)
utils.py:91
Functiontokenize
(prompt)
finetune.py:167
Functiontokenize
(prompt)
finetune_4bit.py:49
Functiontokenize
(prompt)
finetune_fp16.py:175
Functiontokenize
(prompt)
tools/application/chitchat_finetune.py:95
Functiontranspose_matmul_248_kernel
Compute the matrix multiplication C = A x B. A is of shape (M, N) float16 B is of shape (K//8, N) int32 C is of s
tools/quant/quant_linear.py:199
Methodwarmup
(self, *args, **kwargs)
tools/quant/custom_autotune.py:117
Functionwrapper
(*args)
chat.py:365
← previous401–461 of 461, ranked by callers