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
↓ 1 callersMethodpreprocess_gen
(self, data_point)
prompt.py:113
↓ 1 callersMethodprint_loss
(self, name, q_weight, weight_error, timecost)
tools/gptq.py:108
↓ 1 callersFunctionquant_model
(model_path, quant_path, quant_config)
tools/awq/basic_quant.py:7
↓ 1 callersFunctionquantize_row_q4_0
method 5 blocks of QK elements represented with a single float (delta) and QK/2 8-bit ints (i.e QK 4-bit signed integer factors)
tools/Vicuna.cpp/ggml.c:404
↓ 1 callersFunctionquantize_row_q4_1
method 4 blocks of QK elements represented with 2 floats (min + delta) and QK/2 8-bit ints (i.e QK 4-bit unsigned integer factors)
tools/Vicuna.cpp/ggml.c:606
↓ 1 callersMethodrun
(self, *args, **kwargs)
tools/quant/custom_autotune.py:76
↓ 1 callersFunctionrun_eval
Post quantization: Evaluate perplexity on wikitext with EleutherAI Evaluation Harness
tools/awq/eval.py:7
↓ 1 callersFunctionrun_round
(model_path, quant_file, n_generate, input_ids, batch_size, no_safetensors)
tools/awq/benchmark.py:42
↓ 1 callersFunctionsample_top_k
tools/Vicuna.cpp/utils.cpp:338
↓ 1 callersFunctionsched_yield
tools/Vicuna.cpp/ggml.c:68
↓ 1 callersFunctionsilu
(x)
tools/quant/fused_mlp.py:165
↓ 1 callersMethodstream_beam_sample
( self, input_ids, logits_processor, logits_warper, stopping_criteria,
utils.py:383
↓ 1 callersMethodstream_beam_search
( self, generation_config, input_ids, logits_processor, stopping_crite
utils.py:593
↓ 1 callersMethodstream_greedy_search
( self, input_ids, logits_processor, stopping_criteria, generation_con
utils.py:518
↓ 1 callersMethodstream_sample
( self, generation_config, input_ids, logits_processor, logits_warper,
utils.py:306
↓ 1 callersFunctiontranslate_state_dict_key
(k)
tools/merge_lora.py:114
↓ 1 callersFunctionunpermute
(w)
tools/merge_lora.py:108
↓ 1 callersMethodupdate
(self, target_layer)
tools/application/chatglm_lora_finetune.py:181
↓ 1 callersFunctionwarmup
(model)
tools/awq/benchmark.py:10
Method__getitem__
(self, index)
tools/application/chatglm_lora_finetune.py:107
Method__init__
(self, trainer)
finetune_chat.py:215
Method__init__
(self, fmt)
utils.py:49
Method__init__
(self, tokenizer, max_len, add_eos=True)
prompt.py:7
Method__init__
(self, module)
tools/quant_llama.py:51
Method__init__
(self, input_ids)
tools/datautils.py:105
Method__init__
(self, layer, observe=False)
tools/gptq.py:65
Method__init__
( self, gate_proj, down_proj, up_proj, )
tools/quant/fused_mlp.py:173
Method__init__
:param prune_configs_by: a dict of functions that are used to prune configs, fields: 'perf_model': performance model used to predicate running t
tools/quant/custom_autotune.py:16
Method__init__
(self, shape=1)
tools/quant/quantizer.py:9
Method__init__
( self, hidden_size, num_heads, qkv_proj, o_proj, rotary
tools/quant/fused_attn.py:13
Method__init__
(self, bits, groupsize, infeatures, outfeatures, bias)
tools/quant/quant_linear.py:322
Method__init__
(self, in_features, out_features)
tools/application/chatglm_lora_test.py:40
Method__init__
(self, pairs, tokenizer)
tools/application/chatglm_lora_finetune.py:102
Method__len__
(self)
tools/application/chatglm_lora_finetune.py:124
Function__sse_f16x4_load
tools/Vicuna.cpp/ggml.c:1150
Function__sse_f16x4_store
tools/Vicuna.cpp/ggml.c:1161
Function__wasm_f16x4_load
tools/Vicuna.cpp/ggml.c:1040
Function__wasm_f16x4_store
tools/Vicuna.cpp/ggml.c:1051
Method_hook
(args)
tools/quant/custom_autotune.py:36
Method_shape
(self, tensor, seq_len, bsz)
tools/quant/fused_attn.py:33
Functionautotune
Decorator for auto-tuning a :code:`triton.jit`'d function. .. highlight:: python .. code-block:: python @triton.autotune(configs=[ triton.Con
tools/quant/custom_autotune.py:130
Functionautotune_warmup_fused
Pre-tunes the quantized kernel
tools/quant/fused_mlp.py:250
Functionautotune_warmup_linear
Pre-tunes the quantized kernel
tools/quant/quant_linear.py:410
Functioncancel
(history, chatbot)
chat.py:378
Functionclear
()
chat.py:299
Functioncollate_fn
(batch)
tools/application/chatglm_lora_finetune.py:127
Methoddata_collator
(self,)
prompt.py:75
Functiondecorator
(fn)
tools/quant/custom_autotune.py:161
Methodenabled
(self)
tools/quant/quantizer.py:123
Functionevaluate
( inputs, history, temperature=0.1, top_p=0.75, top_k=40, num_beams=4, max_new_tok
chat.py:169
Functionevaluate
( input, temperature=0.1, top_p=0.75, top_k=40, num_beams=4, max_new_tokens=128, m
generate.py:115
Functionevaluate
( input, temperature=0.1, top_p=0.75, top_k=40, num_beams=4, max_new_tokens=128, m
generate_4bit.py:124
Functionevaluate
( input, temperature=0.1, top_p=0.75, top_k=40, nu
tools/quant_generate.py:144
Methodforward
(self, inp, **kwargs)
tools/quant_llama.py:55
Methodforward
(self, x)
tools/quant/fused_mlp.py:197
Methodforward
Input shape: Batch x Time x Channel
tools/quant/fused_attn.py:36
Methodforward
(ctx, input, qweight, scales, qzeros, g_idx, bits, maxq, no_group)
tools/quant/quant_linear.py:302
Methodforward
(self, x)
tools/quant/quant_linear.py:390
Methodforward
(self, x)
tools/application/chatglm_lora_test.py:56
Methodforward
(self, x)
tools/application/chatglm_lora_finetune.py:191
Functionfrom_json
(path)
utils.py:96
Functionfrom_jsonl
(path)
utils.py:99
Functionfusedmatmul_248_kernel
Computes: C = silu(A * B1) * (A * B2) A is of shape (M, K) float16 B is of shape (K//8, N) int32 C is of shape (M
tools/quant/fused_mlp.py:79
Functiongenerate_and_tokenize_prompt
(data_point)
finetune.py:182
Functiongenerate_and_tokenize_prompt
(data_point)
finetune_4bit.py:64
Functiongenerate_and_tokenize_prompt
(data_point)
finetune_fp16.py:190
Functiongenerate_and_tokenize_prompt
(data_point)
tools/application/chitchat_finetune.py:108
Functiongenerate_prompt
(data_point)
finetune.py:144
Functiongenerate_prompt
(data_point)
finetune_4bit.py:27
Functiongenerate_prompt
(data_point)
finetune_fp16.py:152
Methodget_data_collator
()
prompt.py:208
Functionget_peft_state_maybe_zero_3
(state_dict, bias)
finetune_fp16.py:25
Functionggml_abs
tools/Vicuna.cpp/ggml.c:3371
Functionggml_abs_inplace
tools/Vicuna.cpp/ggml.c:3377
Functionggml_add_inplace
tools/Vicuna.cpp/ggml.c:3077
Functionggml_compute_fp16_to_fp32
tools/Vicuna.cpp/ggml.c:187
Functionggml_compute_fp32_to_fp16
tools/Vicuna.cpp/ggml.c:210
Functionggml_conv_1d_1s
tools/Vicuna.cpp/ggml.c:4096
Functionggml_conv_1d_2s
tools/Vicuna.cpp/ggml.c:4123
Functionggml_cpy_inplace
tools/Vicuna.cpp/ggml.c:3758
Functionggml_div_inplace
tools/Vicuna.cpp/ggml.c:3202
Functionggml_dup
tools/Vicuna.cpp/ggml.c:3033
Functionggml_dup_inplace
tools/Vicuna.cpp/ggml.c:3039
Functionggml_flash_attn
tools/Vicuna.cpp/ggml.c:4150
Functionggml_flash_ff
tools/Vicuna.cpp/ggml.c:4181
Functionggml_gelu
tools/Vicuna.cpp/ggml.c:3542
Functionggml_gelu_inplace
tools/Vicuna.cpp/ggml.c:3548
Functionggml_get_data_f32
tools/Vicuna.cpp/ggml.c:2998
Functionggml_graph_compute_thread
tools/Vicuna.cpp/ggml.c:9185
Functionggml_lookup_fp16_to_fp32
tools/Vicuna.cpp/ggml.c:263
Functionggml_mean
tools/Vicuna.cpp/ggml.c:3300
Functionggml_mul_inplace
tools/Vicuna.cpp/ggml.c:3159
Functionggml_neg
tools/Vicuna.cpp/ggml.c:3440
Functionggml_neg_inplace
tools/Vicuna.cpp/ggml.c:3446
Functionggml_new_tensor_4d
tools/Vicuna.cpp/ggml.c:2656
Functionggml_norm
tools/Vicuna.cpp/ggml.c:3611
Functionggml_norm_inplace
tools/Vicuna.cpp/ggml.c:3617
Functionggml_opt
tools/Vicuna.cpp/ggml.c:10493
Functionggml_opt_default_params
tools/Vicuna.cpp/ggml.c:10432
Functionggml_print_objects
tools/Vicuna.cpp/ggml.c:2238
← previousnext →301–400 of 461, ranked by callers