MCPcopy Create free account

hub / github.com/microsoft/BitNet / functions

Functions397 in github.com/microsoft/BitNet

Method__init__
(self, tokenizer: Tokenizer)
gpu/tokenizer.py:198
Method__init__
( self, args: GenArgs, model_args: fast.ModelArgs, prefill_model: fast.Tra
gpu/generate.py:79
Method__init__
(self, in_features: int, out_features: int, bias: bool = False)
gpu/model.py:60
Method__init__
( self, dim: int, hidden_dim: int, norm_eps: float, use_kernel: b
gpu/model.py:169
Method__init__
(self, args: ModelArgs)
gpu/model.py:201
Method__init__
(self, args: ModelArgs)
gpu/model.py:247
Method__init__
(self)
gpu/stats.py:32
Method__post_init__
(self)
gpu/convert_safetensors.py:26
Method__repr__
(self)
utils/convert-ms-to-gguf-bitnet.py:404
Method__repr__
(self)
utils/convert-ms-to-gguf-bitnet.py:490
Method__repr__
(self)
utils/convert-ms-to-gguf-bitnet.py:561
Method__repr__
(self)
utils/convert-ms-to-gguf-bitnet.py:684
Method__repr__
(self)
utils/convert.py:404
Method__repr__
(self)
utils/convert.py:490
Method__repr__
(self)
utils/convert.py:561
Method__repr__
(self)
utils/convert.py:684
Methodall_tokens
(self)
utils/convert-ms-to-gguf-bitnet.py:486
Methodall_tokens
(self)
utils/convert-ms-to-gguf-bitnet.py:557
Methodall_tokens
(self)
utils/convert-ms-to-gguf-bitnet.py:680
Methodall_tokens
(self)
utils/convert.py:486
Methodall_tokens
(self)
utils/convert.py:557
Methodastype
(self, data_type: DataType)
utils/convert-ms-to-gguf-bitnet.py:788
Methodastype
(self, data_type: DataType)
utils/convert-ms-to-gguf-bitnet.py:844
Methodastype
(self, data_type: DataType)
utils/convert.py:788
Methodastype
(self, data_type: DataType)
utils/convert.py:844
Functioncache_prefix
Take a prefix view of a larger cache. The original cache object remains of identical size and valid after the shrinked alias has bee
gpu/model.py:346
Methoddecode
(self, tokens: List[int])
gpu/tokenizer.py:202
Methoddo_item
(item: tuple[str, LazyTensor])
utils/convert-ms-to-gguf-bitnet.py:1301
Methoddo_item
(item: tuple[str, LazyTensor])
utils/convert.py:1232
Methodelements_to_bytes
(self, n_elements: int)
utils/convert-ms-to-gguf-bitnet.py:87
Methodelements_to_bytes
(self, n_elements: int)
utils/convert.py:87
Methodforward
(self, input)
gpu/model.py:73
Methodforward
( self, x: torch.Tensor, cache: LayerCache, attn_bias: AttnBias, )
gpu/model.py:121
Methodforward
(self, x: torch.Tensor)
gpu/model.py:192
Methodforward
( self, x: torch.Tensor, cache: LayerCache, attn_bias: AttnBias, )
gpu/model.py:231
Methodforward
( self, token_values: torch.Tensor, token_lengths: torch.Tensor, start_pos
gpu/model.py:283
Methodfunc
(modelcls: type[Model])
utils/convert-hf-to-gguf-bitnet.py:229
Methodfunc
(modelcls: type[Model])
utils/generate-dummy-bitnet-model.py:301
Methodgenerate_extra_tensors
(self)
utils/convert-hf-to-gguf-bitnet.py:916
Methodget_ppl_value
(result)
utils/test_perplexity.py:502
Methodget_vocab_base
(self)
utils/generate-dummy-bitnet-model.py:328
Methodhas_newline_token
(self)
utils/convert-ms-to-gguf-bitnet.py:677
Methodhas_newline_token
(self)
utils/convert.py:677
Functionload
()
utils/convert-ms-to-gguf-bitnet.py:892
Functionload
()
utils/convert.py:892
Methodload
(self)
utils/convert-ms-to-gguf-bitnet.py:837
Methodload
(self)
utils/convert.py:837
Functionmain
(ckpt_dir: str, interactive: bool = False, chat_format: bool = False, sampling: bool = False)
gpu/generate.py:322
Functionmake_cache
Allocate a cache to be used with the Transformer module. Args: args (ModelArgs): the model configuration. length (int):
gpu/model.py:299
Methodmaybe_do_quantize
(item: tuple[DataType, NDArray])
utils/convert-ms-to-gguf-bitnet.py:1307
Methodmaybe_do_quantize
(item: tuple[DataType, NDArray])
utils/convert.py:1238
Methodmodel_arch
(self)
utils/convert-hf-to-gguf-bitnet.py:68
Methodmodel_arch
(self)
utils/generate-dummy-bitnet-model.py:140
Functionmust_read
(fp: IO[bytes], length: int)
utils/convert-ms-to-gguf-bitnet.py:1053
Functionmust_read
(fp: IO[bytes], length: int)
utils/convert.py:979
Functionpack_experts_lazy
(lazy_tensors: list[LazyTensor])
utils/convert-ms-to-gguf-bitnet.py:1020
Methodpart
(self, n_part: int)
utils/convert-ms-to-gguf-bitnet.py:803
Methodpart
(self, n_part: int)
utils/convert.py:803
Functionpart_lazy
(lazy_tensor: LazyTensor, n_part: int)
utils/convert-ms-to-gguf-bitnet.py:937
Functionpart_lazy_k
(lazy_tensor: LazyTensor, n_part: int)
utils/convert-ms-to-gguf-bitnet.py:968
Functionpart_lazy_q
(lazy_tensor: LazyTensor, n_part: int)
utils/convert-ms-to-gguf-bitnet.py:960
Functionpart_lazy_rope
(lazy_tensor: LazyTensor)
utils/convert-ms-to-gguf-bitnet.py:1004
Functionpart_lazy_v
(lazy_tensor: LazyTensor, n_part: int)
utils/convert-ms-to-gguf-bitnet.py:976
Functionpart_lazy_w1
(lazy_tensor: LazyTensor, n_part: int)
utils/convert-ms-to-gguf-bitnet.py:986
Functionpart_lazy_w3
(lazy_tensor: LazyTensor, n_part: int)
utils/convert-ms-to-gguf-bitnet.py:995
Functionpart_lazy_weight_quant
(lazy_tensor: LazyTensor, name)
utils/convert-ms-to-gguf-bitnet.py:1011
Methodpermute
(self, n_head: int, n_head_kv: int)
utils/convert-ms-to-gguf-bitnet.py:807
Methodpermute
(self, n_head: int, n_head_kv: int)
utils/convert.py:807
Functionpermute_lazy
(lazy_tensor: LazyTensor, n_head: int, n_head_kv: int)
utils/convert-ms-to-gguf-bitnet.py:924
Methodpermute_part
(self, n_part: int, n_head: int, n_head_kv: int)
utils/convert-ms-to-gguf-bitnet.py:799
Methodpermute_part
(self, n_part: int, n_head: int, n_head_kv: int)
utils/convert.py:799
Functionpermute_part_lazy
(lazy_tensor: LazyTensor, n_part: int, n_head: int, n_head_kv: int)
utils/convert-ms-to-gguf-bitnet.py:930
Methodprepare_tensors
(self)
utils/convert-hf-to-gguf-bitnet.py:945
Methodquantize
(self, arr: NDArray)
utils/convert-ms-to-gguf-bitnet.py:95
Methodquantize
(self, arr: NDArray)
utils/convert.py:95
Methodquantize_blocks_q8_0
(blocks: NDArray)
utils/convert-ms-to-gguf-bitnet.py:102
Methodquantize_blocks_q8_0
(blocks: NDArray)
utils/convert.py:102
Functionread_gguf_file
Reads and prints key-value pairs and tensor information from a GGUF file in an improved format. Parameters: - gguf_file_path: Path to th
utils/generate-dummy-bitnet-model.py:990
Functionread_model_config
(model_dir: str)
utils/convert-hf-to-gguf-bitnet.py:675
Methodregister
(cls, *names: str)
utils/generate-dummy-bitnet-model.py:298
Methodreplay
(tokens, seq_lens=None)
gpu/generate.py:147
Methodset_gguf_parameters
(self)
utils/convert-hf-to-gguf-bitnet.py:834
Methodset_gguf_parameters
(self)
utils/convert-hf-to-gguf-bitnet.py:962
Methodset_gguf_parameters
(self)
utils/generate-dummy-bitnet-model.py:796
Methodset_vocab
(self)
utils/convert-hf-to-gguf-bitnet.py:686
Methodset_vocab
(self)
utils/convert-hf-to-gguf-bitnet.py:959
Methodset_vocab
(self)
utils/generate-dummy-bitnet-model.py:793
Functionsignal_handler
(sig, frame)
setup_env.py:235
Functionsignal_handler
(sig, frame)
run_inference_server.py:46
Functionsignal_handler
(sig, frame)
run_inference.py:39
Methodto_ggml
(self)
utils/convert-ms-to-gguf-bitnet.py:796
Methodto_ggml
(self)
utils/convert.py:796
Functiontop_p
Perform top-p (nucleus) sampling on a probability distribution. Args: probs (torch.Tensor): probability distribution tensor.
gpu/sample_utils.py:9
Methodtrim_answer
Trim the answer to end it on an eos token.
gpu/generate.py:290
Methodwrite_tensors
(self)
utils/convert-hf-to-gguf-bitnet.py:708
Methodwrite_tensors
(self)
utils/convert-hf-to-gguf-bitnet.py:986
Methodwrite_tensors
(self)
utils/generate-dummy-bitnet-model.py:861
← previous301–397 of 397, ranked by callers