From 65743ba1b98eff7837397e4b7386e3d0aeb80ed7 Mon Sep 17 00:00:00 2001 From: Meng Zhang Date: Sat, 18 Mar 2023 22:19:07 +0800 Subject: [PATCH] Remove useless --- converter/huggingface_gptj_convert.py | 253 ------------------ converter/testdata/1-gpu/config.ini | 86 ------ .../1-gpu/model.final_layernorm.bias.bin | 3 - .../1-gpu/model.final_layernorm.weight.bin | 3 - ...odel.layers.0.attention.dense.weight.0.bin | 3 - ...s.0.attention.query_key_value.weight.0.bin | 3 - .../model.layers.0.input_layernorm.bias.bin | 3 - .../model.layers.0.input_layernorm.weight.bin | 3 - .../model.layers.0.mlp.dense_4h_to_h.bias.bin | 3 - ...el.layers.0.mlp.dense_4h_to_h.weight.0.bin | 3 - ...odel.layers.0.mlp.dense_h_to_4h.bias.0.bin | 3 - ...el.layers.0.mlp.dense_h_to_4h.weight.0.bin | 3 - ...odel.layers.1.attention.dense.weight.0.bin | 3 - ...s.1.attention.query_key_value.weight.0.bin | 3 - .../model.layers.1.input_layernorm.bias.bin | 3 - .../model.layers.1.input_layernorm.weight.bin | 3 - .../model.layers.1.mlp.dense_4h_to_h.bias.bin | 3 - ...el.layers.1.mlp.dense_4h_to_h.weight.0.bin | 3 - ...odel.layers.1.mlp.dense_h_to_4h.bias.0.bin | 3 - ...el.layers.1.mlp.dense_h_to_4h.weight.0.bin | 3 - ...odel.layers.2.attention.dense.weight.0.bin | 3 - ...s.2.attention.query_key_value.weight.0.bin | 3 - .../model.layers.2.input_layernorm.bias.bin | 3 - .../model.layers.2.input_layernorm.weight.bin | 3 - .../model.layers.2.mlp.dense_4h_to_h.bias.bin | 3 - ...el.layers.2.mlp.dense_4h_to_h.weight.0.bin | 3 - ...odel.layers.2.mlp.dense_h_to_4h.bias.0.bin | 3 - ...el.layers.2.mlp.dense_h_to_4h.weight.0.bin | 3 - ...odel.layers.3.attention.dense.weight.0.bin | 3 - ...s.3.attention.query_key_value.weight.0.bin | 3 - .../model.layers.3.input_layernorm.bias.bin | 3 - .../model.layers.3.input_layernorm.weight.bin | 3 - .../model.layers.3.mlp.dense_4h_to_h.bias.bin | 3 - ...el.layers.3.mlp.dense_4h_to_h.weight.0.bin | 3 - ...odel.layers.3.mlp.dense_h_to_4h.bias.0.bin | 3 - ...el.layers.3.mlp.dense_h_to_4h.weight.0.bin | 3 - ...odel.layers.4.attention.dense.weight.0.bin | 3 - ...s.4.attention.query_key_value.weight.0.bin | 3 - .../model.layers.4.input_layernorm.bias.bin | 3 - .../model.layers.4.input_layernorm.weight.bin | 3 - .../model.layers.4.mlp.dense_4h_to_h.bias.bin | 3 - ...el.layers.4.mlp.dense_4h_to_h.weight.0.bin | 3 - ...odel.layers.4.mlp.dense_h_to_4h.bias.0.bin | 3 - ...el.layers.4.mlp.dense_h_to_4h.weight.0.bin | 3 - .../testdata/1-gpu/model.lm_head.bias.bin | 3 - .../testdata/1-gpu/model.lm_head.weight.bin | 3 - converter/testdata/1-gpu/model.wte.bin | 3 - 47 files changed, 474 deletions(-) delete mode 100644 converter/huggingface_gptj_convert.py delete mode 100644 converter/testdata/1-gpu/config.ini delete mode 100644 converter/testdata/1-gpu/model.final_layernorm.bias.bin delete mode 100644 converter/testdata/1-gpu/model.final_layernorm.weight.bin delete mode 100644 converter/testdata/1-gpu/model.layers.0.attention.dense.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.0.attention.query_key_value.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.0.input_layernorm.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.0.input_layernorm.weight.bin delete mode 100644 converter/testdata/1-gpu/model.layers.0.mlp.dense_4h_to_h.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.0.mlp.dense_4h_to_h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.0.mlp.dense_h_to_4h.bias.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.0.mlp.dense_h_to_4h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.1.attention.dense.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.1.attention.query_key_value.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.1.input_layernorm.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.1.input_layernorm.weight.bin delete mode 100644 converter/testdata/1-gpu/model.layers.1.mlp.dense_4h_to_h.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.1.mlp.dense_4h_to_h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.1.mlp.dense_h_to_4h.bias.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.1.mlp.dense_h_to_4h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.2.attention.dense.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.2.attention.query_key_value.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.2.input_layernorm.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.2.input_layernorm.weight.bin delete mode 100644 converter/testdata/1-gpu/model.layers.2.mlp.dense_4h_to_h.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.2.mlp.dense_4h_to_h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.2.mlp.dense_h_to_4h.bias.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.2.mlp.dense_h_to_4h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.3.attention.dense.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.3.attention.query_key_value.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.3.input_layernorm.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.3.input_layernorm.weight.bin delete mode 100644 converter/testdata/1-gpu/model.layers.3.mlp.dense_4h_to_h.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.3.mlp.dense_4h_to_h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.3.mlp.dense_h_to_4h.bias.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.3.mlp.dense_h_to_4h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.4.attention.dense.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.4.attention.query_key_value.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.4.input_layernorm.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.4.input_layernorm.weight.bin delete mode 100644 converter/testdata/1-gpu/model.layers.4.mlp.dense_4h_to_h.bias.bin delete mode 100644 converter/testdata/1-gpu/model.layers.4.mlp.dense_4h_to_h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.4.mlp.dense_h_to_4h.bias.0.bin delete mode 100644 converter/testdata/1-gpu/model.layers.4.mlp.dense_h_to_4h.weight.0.bin delete mode 100644 converter/testdata/1-gpu/model.lm_head.bias.bin delete mode 100644 converter/testdata/1-gpu/model.lm_head.weight.bin delete mode 100644 converter/testdata/1-gpu/model.wte.bin diff --git a/converter/huggingface_gptj_convert.py b/converter/huggingface_gptj_convert.py deleted file mode 100644 index c435819..0000000 --- a/converter/huggingface_gptj_convert.py +++ /dev/null @@ -1,253 +0,0 @@ -# Copyright (c) 2021-2022, NVIDIA CORPORATION. All rights reserved. -# Modified by Brendan Dolan-Gavitt, 2022 -# Modified by Meng Zhang, 2023 -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -import argparse -import configparser -import multiprocessing -import os -import sys -from pathlib import Path - -import numpy as np -import torch -from transformers import GPTJForCausalLM - -dir_path = os.path.dirname(os.path.realpath(__file__)) -sys.path.append(dir_path + "/../../../..") -sys.path.append(dir_path) - - -def get_weight_data_type(data_type): - if data_type == "fp32": - return np.float32 - elif data_type == "fp16": - return np.float16 - else: - assert False, f"Invalid weight data type {data_type}" - - -def split_and_convert_process(i, saved_dir, factor, key, val): - if ( - key.find("input_layernorm.weight") != -1 - or key.find("input_layernorm.bias") != -1 - or key.find("attention.dense.bias") != -1 - or key.find("post_attention_layernorm.weight") != -1 - or key.find("post_attention_layernorm.bias") != -1 - or key.find("mlp.dense_4h_to_h.bias") != -1 - or key.find("final_layernorm.weight") != -1 - or key.find("final_layernorm.bias") != -1 - ): - - # shared weights, only need to convert the weights of rank 0 - if i == 0: - saved_path = saved_dir + "/model." + key + ".bin" - val.tofile(saved_path) - - elif ( - key.find("attention.dense.weight") != -1 - or key.find("mlp.dense_4h_to_h.weight") != -1 - ): - split_vals = np.split(val, factor, axis=0) - for j in range(factor): - saved_path = saved_dir + "/model." + key + ".%d.bin" % (i * factor + j) - split_vals[j].tofile(saved_path) - - elif ( - key.find("mlp.dense_h_to_4h.weight") != -1 - or key.find("mlp.dense_h_to_4h.bias") != -1 - ): - - split_vals = np.split(val, factor, axis=-1) - for j in range(factor): - saved_path = saved_dir + "/model." + key + ".%d.bin" % (i * factor + j) - split_vals[j].tofile(saved_path) - - elif key.find("attention.query_key_value.weight") != -1: - split_vals = np.split(val, factor, axis=-1) - - for j in range(factor): - saved_path = saved_dir + "/model." + key + ".%d.bin" % (i * factor + j) - split_vals[j].tofile(saved_path) - - else: - print("[ERROR] cannot find key '{}'".format(key)) - - -def split_and_convert(args): - saved_dir = args.saved_dir + "/%d-gpu/" % args.infer_gpu_num - - if os.path.exists(saved_dir) is False: - os.makedirs(saved_dir) - - t_gpu_num = args.trained_gpu_num - i_gpu_num = args.infer_gpu_num - assert i_gpu_num % t_gpu_num == 0 - - factor = (int)(i_gpu_num / t_gpu_num) - - model = GPTJForCausalLM.from_pretrained(args.in_file) - - try: - config = configparser.ConfigParser() - config["gpt"] = {} - for key in vars(args): - config["gpt"][key] = f"{vars(args)[key]}" - for k, v in vars(model.config).items(): - config["gpt"][k] = f"{v}" - config["gpt"]["weight_data_type"] = args.weight_data_type - with open((Path(saved_dir) / "config.ini").as_posix(), "w") as configfile: - config.write(configfile) - except Exception: - print("Fail to save the config in config.ini.") - np_weight_data_type = get_weight_data_type(args.weight_data_type) - - huggingface_model_name_pattern = [ - "ln_1.bias", - "ln_1.weight", - "attn.q_proj.weight", - "attn.out_proj.weight", - "mlp.fc_in.bias", - "mlp.fc_in.weight", - "mlp.fc_out.bias", - "mlp.fc_out.weight", - ] - - ft_model_name_pattern = [ - "input_layernorm.bias", - "input_layernorm.weight", - "attention.query_key_value.weight", - "attention.dense.weight", - "mlp.dense_h_to_4h.bias", - "mlp.dense_h_to_4h.weight", - "mlp.dense_4h_to_h.bias", - "mlp.dense_4h_to_h.weight", - ] - - torch.multiprocessing.set_start_method("spawn") - pool = multiprocessing.Pool(args.processes) - for name, param in model.named_parameters(): - if name.find("weight") == -1 and name.find("bias") == -1: - continue - print(name) - if name == "transformer.wte.weight": - param.detach().cpu().numpy().astype(np_weight_data_type).tofile( - saved_dir + "model.wte.bin" - ) - elif name == "transformer.ln_f.bias": - param.detach().cpu().numpy().astype(np_weight_data_type).tofile( - saved_dir + "model.final_layernorm.bias.bin" - ) - elif name == "transformer.ln_f.weight": - param.detach().cpu().numpy().astype(np_weight_data_type).tofile( - saved_dir + "model.final_layernorm.weight.bin" - ) - elif name == "lm_head.weight": - param.detach().cpu().numpy().astype(np_weight_data_type).tofile( - saved_dir + "model.lm_head.weight.bin" - ) - elif name == "lm_head.bias": - param.detach().cpu().numpy().astype(np_weight_data_type).tofile( - saved_dir + "model.lm_head.bias.bin" - ) - else: - for i in range(len(huggingface_model_name_pattern)): - if name.find(huggingface_model_name_pattern[i]) != -1: - # Special case for QKV weights - if name.find("attn.q_proj.weight") != -1: - layer = name.split(".")[2] - base_k = f"transformer.h.{layer}." - w = model.state_dict() - QKV_w = torch.stack( - [ - w[base_k + "attn.q_proj.weight"], - w[base_k + "attn.k_proj.weight"], - w[base_k + "attn.v_proj.weight"], - ] - ) # [qkv, n_heads * dim_head, latent_space] - QKV_w = QKV_w.permute(2, 0, 1) - weights = ( - QKV_w.detach().cpu().numpy().astype(np_weight_data_type) - ) - else: - weights = ( - param.detach().cpu().numpy().astype(np_weight_data_type) - ) - - # Some weights need to be transposed - if ( - name.find("mlp.fc_in.weight") != -1 - or name.find("mlp.fc_out.weight") != -1 - or name.find("attn.out_proj.weight") != -1 - ): - weights = weights.T - - new_name = name.replace("transformer.h.", "layers.").replace( - huggingface_model_name_pattern[i], ft_model_name_pattern[i] - ) - - pool.starmap( - split_and_convert_process, - [(0, saved_dir, factor, new_name, weights)], - ) - - pool.close() - pool.join() - - -if __name__ == "__main__": - parser = argparse.ArgumentParser(formatter_class=argparse.RawTextHelpFormatter) - parser.add_argument( - "-saved_dir", "-o", type=str, help="file name of output file", required=True - ) - parser.add_argument( - "-in_file", "-i", type=str, help="HF model name or directory", required=True - ) - parser.add_argument( - "-trained_gpu_num", - "-t_g", - type=int, - help="How many gpus for training", - default=1, - ) - parser.add_argument( - "-infer_gpu_num", - "-i_g", - type=int, - help="How many gpus for inference", - required=True, - ) - parser.add_argument( - "-processes", - "-p", - type=int, - help="How many processes to spawn for conversion (default: 4)", - default=4, - ) - parser.add_argument( - "-weight_data_type", - type=str, - default="fp32", - choices=["fp32", "fp16"], - help="output weight data type", - ) - - args = parser.parse_args() - print("\n=============== Argument ===============") - for key in vars(args): - print("{}: {}".format(key, vars(args)[key])) - print("========================================") - - split_and_convert(args) diff --git a/converter/testdata/1-gpu/config.ini b/converter/testdata/1-gpu/config.ini deleted file mode 100644 index 975a385..0000000 --- a/converter/testdata/1-gpu/config.ini +++ /dev/null @@ -1,86 +0,0 @@ -[gpt] -saved_dir = out -in_file = hf-internal-testing/tiny-random-gptj -trained_gpu_num = 1 -infer_gpu_num = 1 -processes = 4 -weight_data_type = fp32 -vocab_size = 1000 -n_positions = 512 -n_embd = 32 -n_layer = 5 -n_head = 4 -n_inner = None -rotary_dim = 4 -activation_function = gelu_new -resid_pdrop = 0.0 -embd_pdrop = 0.0 -attn_pdrop = 0.0 -layer_norm_epsilon = 1e-05 -initializer_range = 0.02 -use_cache = True -bos_token_id = 98 -eos_token_id = 98 -return_dict = True -output_hidden_states = False -output_attentions = False -torchscript = False -torch_dtype = None -use_bfloat16 = False -tf_legacy_loss = False -pruned_heads = {} -tie_word_embeddings = False -is_encoder_decoder = False -is_decoder = False -cross_attention_hidden_size = None -add_cross_attention = False -tie_encoder_decoder = False -max_length = 20 -min_length = 0 -do_sample = False -early_stopping = False -num_beams = 1 -num_beam_groups = 1 -diversity_penalty = 0.0 -temperature = 1.0 -top_k = 50 -top_p = 1.0 -typical_p = 1.0 -repetition_penalty = 1.0 -length_penalty = 1.0 -no_repeat_ngram_size = 0 -encoder_no_repeat_ngram_size = 0 -bad_words_ids = None -num_return_sequences = 1 -chunk_size_feed_forward = 0 -output_scores = False -return_dict_in_generate = False -forced_bos_token_id = None -forced_eos_token_id = None -remove_invalid_values = False -exponential_decay_length_penalty = None -suppress_tokens = None -begin_suppress_tokens = None -architectures = None -finetuning_task = None -id2label = {0: 'LABEL_0', 1: 'LABEL_1'} -label2id = {'LABEL_0': 0, 'LABEL_1': 1} -tokenizer_class = None -prefix = None -pad_token_id = 98 -sep_token_id = None -decoder_start_token_id = None -task_specific_params = None -problem_type = None -_name_or_path = hf-internal-testing/tiny-random-gptj -_commit_hash = b96595a4bcdeb272096214589efa0314259853a0 -transformers_version = 4.11.0.dev0 -attention_probs_dropout_prob = 0.0 -gradient_checkpointing = False -hidden_act = gelu -hidden_dropout_prob = 0.0 -intermediate_size = 37 -model_type = gptj -n_ctx = 512 -scale_attn_weights = True -type_vocab_size = 16 diff --git a/converter/testdata/1-gpu/model.final_layernorm.bias.bin b/converter/testdata/1-gpu/model.final_layernorm.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.final_layernorm.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.final_layernorm.weight.bin b/converter/testdata/1-gpu/model.final_layernorm.weight.bin deleted file mode 100644 index d0c1df9..0000000 --- a/converter/testdata/1-gpu/model.final_layernorm.weight.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:b638277a8690e175a9137feff1e43c067f9faf4e2f600caf468fb05b0403b717 -size 128 diff --git a/converter/testdata/1-gpu/model.layers.0.attention.dense.weight.0.bin b/converter/testdata/1-gpu/model.layers.0.attention.dense.weight.0.bin deleted file mode 100644 index 48a4419..0000000 --- a/converter/testdata/1-gpu/model.layers.0.attention.dense.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:125bd07dc5e7a5c6444f7689ba78df0f4c7959b6dcfbaf7c89edd0634a147ea0 -size 4096 diff --git a/converter/testdata/1-gpu/model.layers.0.attention.query_key_value.weight.0.bin b/converter/testdata/1-gpu/model.layers.0.attention.query_key_value.weight.0.bin deleted file mode 100644 index 33c2eca..0000000 --- a/converter/testdata/1-gpu/model.layers.0.attention.query_key_value.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:11fd52f2b94fad5fe54e2e03c5848050703ec4d798a43a4c1813f109a6703883 -size 12288 diff --git a/converter/testdata/1-gpu/model.layers.0.input_layernorm.bias.bin b/converter/testdata/1-gpu/model.layers.0.input_layernorm.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.0.input_layernorm.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.0.input_layernorm.weight.bin b/converter/testdata/1-gpu/model.layers.0.input_layernorm.weight.bin deleted file mode 100644 index d0c1df9..0000000 --- a/converter/testdata/1-gpu/model.layers.0.input_layernorm.weight.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:b638277a8690e175a9137feff1e43c067f9faf4e2f600caf468fb05b0403b717 -size 128 diff --git a/converter/testdata/1-gpu/model.layers.0.mlp.dense_4h_to_h.bias.bin b/converter/testdata/1-gpu/model.layers.0.mlp.dense_4h_to_h.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.0.mlp.dense_4h_to_h.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.0.mlp.dense_4h_to_h.weight.0.bin b/converter/testdata/1-gpu/model.layers.0.mlp.dense_4h_to_h.weight.0.bin deleted file mode 100644 index 4da1133..0000000 --- a/converter/testdata/1-gpu/model.layers.0.mlp.dense_4h_to_h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c7459667751237cef87ba33e5faf46e82ea33f7f53aaae7bf95a8667b6f9e639 -size 16384 diff --git a/converter/testdata/1-gpu/model.layers.0.mlp.dense_h_to_4h.bias.0.bin b/converter/testdata/1-gpu/model.layers.0.mlp.dense_h_to_4h.bias.0.bin deleted file mode 100644 index c30ec76..0000000 --- a/converter/testdata/1-gpu/model.layers.0.mlp.dense_h_to_4h.bias.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:076a27c79e5ace2a3d47f9dd2e83e4ff6ea8872b3c2218f66c92b89b55f36560 -size 512 diff --git a/converter/testdata/1-gpu/model.layers.0.mlp.dense_h_to_4h.weight.0.bin b/converter/testdata/1-gpu/model.layers.0.mlp.dense_h_to_4h.weight.0.bin deleted file mode 100644 index 668eaee..0000000 --- a/converter/testdata/1-gpu/model.layers.0.mlp.dense_h_to_4h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:55de52e2d73e7e92962bc02673a7dce6123d77e486ac4f8ce7bead1a6727d227 -size 16384 diff --git a/converter/testdata/1-gpu/model.layers.1.attention.dense.weight.0.bin b/converter/testdata/1-gpu/model.layers.1.attention.dense.weight.0.bin deleted file mode 100644 index 97a8f58..0000000 --- a/converter/testdata/1-gpu/model.layers.1.attention.dense.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:ec12f3d7c73cbc440b476ad66bb4a1ef43881c4d740e8114dc1578bb0197a17e -size 4096 diff --git a/converter/testdata/1-gpu/model.layers.1.attention.query_key_value.weight.0.bin b/converter/testdata/1-gpu/model.layers.1.attention.query_key_value.weight.0.bin deleted file mode 100644 index 33b977e..0000000 --- a/converter/testdata/1-gpu/model.layers.1.attention.query_key_value.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:86d3c23c240260084ac27bd98d52524f0b3559d8106839d85bc927b44750bd81 -size 12288 diff --git a/converter/testdata/1-gpu/model.layers.1.input_layernorm.bias.bin b/converter/testdata/1-gpu/model.layers.1.input_layernorm.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.1.input_layernorm.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.1.input_layernorm.weight.bin b/converter/testdata/1-gpu/model.layers.1.input_layernorm.weight.bin deleted file mode 100644 index d0c1df9..0000000 --- a/converter/testdata/1-gpu/model.layers.1.input_layernorm.weight.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:b638277a8690e175a9137feff1e43c067f9faf4e2f600caf468fb05b0403b717 -size 128 diff --git a/converter/testdata/1-gpu/model.layers.1.mlp.dense_4h_to_h.bias.bin b/converter/testdata/1-gpu/model.layers.1.mlp.dense_4h_to_h.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.1.mlp.dense_4h_to_h.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.1.mlp.dense_4h_to_h.weight.0.bin b/converter/testdata/1-gpu/model.layers.1.mlp.dense_4h_to_h.weight.0.bin deleted file mode 100644 index d6e7188..0000000 --- a/converter/testdata/1-gpu/model.layers.1.mlp.dense_4h_to_h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:75338924b49fb652d556c260013520b58ca70c8bd782bf59732c5ca8d1de111d -size 16384 diff --git a/converter/testdata/1-gpu/model.layers.1.mlp.dense_h_to_4h.bias.0.bin b/converter/testdata/1-gpu/model.layers.1.mlp.dense_h_to_4h.bias.0.bin deleted file mode 100644 index c30ec76..0000000 --- a/converter/testdata/1-gpu/model.layers.1.mlp.dense_h_to_4h.bias.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:076a27c79e5ace2a3d47f9dd2e83e4ff6ea8872b3c2218f66c92b89b55f36560 -size 512 diff --git a/converter/testdata/1-gpu/model.layers.1.mlp.dense_h_to_4h.weight.0.bin b/converter/testdata/1-gpu/model.layers.1.mlp.dense_h_to_4h.weight.0.bin deleted file mode 100644 index 622a346..0000000 --- a/converter/testdata/1-gpu/model.layers.1.mlp.dense_h_to_4h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:be97681ddf43d42ee25d0439c6958edac3549b0edda8a4e3e8bce4275916bb7c -size 16384 diff --git a/converter/testdata/1-gpu/model.layers.2.attention.dense.weight.0.bin b/converter/testdata/1-gpu/model.layers.2.attention.dense.weight.0.bin deleted file mode 100644 index 9fef556..0000000 --- a/converter/testdata/1-gpu/model.layers.2.attention.dense.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:52bfbb09f85dba050a997ddd1869ae74aa8e76a70a63fd4b7c047531128bec07 -size 4096 diff --git a/converter/testdata/1-gpu/model.layers.2.attention.query_key_value.weight.0.bin b/converter/testdata/1-gpu/model.layers.2.attention.query_key_value.weight.0.bin deleted file mode 100644 index f84cc8b..0000000 --- a/converter/testdata/1-gpu/model.layers.2.attention.query_key_value.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:869b24c7a494832730ba41e0286966b23e700e4239d39198e78551a7ef5e8d7f -size 12288 diff --git a/converter/testdata/1-gpu/model.layers.2.input_layernorm.bias.bin b/converter/testdata/1-gpu/model.layers.2.input_layernorm.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.2.input_layernorm.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.2.input_layernorm.weight.bin b/converter/testdata/1-gpu/model.layers.2.input_layernorm.weight.bin deleted file mode 100644 index d0c1df9..0000000 --- a/converter/testdata/1-gpu/model.layers.2.input_layernorm.weight.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:b638277a8690e175a9137feff1e43c067f9faf4e2f600caf468fb05b0403b717 -size 128 diff --git a/converter/testdata/1-gpu/model.layers.2.mlp.dense_4h_to_h.bias.bin b/converter/testdata/1-gpu/model.layers.2.mlp.dense_4h_to_h.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.2.mlp.dense_4h_to_h.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.2.mlp.dense_4h_to_h.weight.0.bin b/converter/testdata/1-gpu/model.layers.2.mlp.dense_4h_to_h.weight.0.bin deleted file mode 100644 index a71148a..0000000 --- a/converter/testdata/1-gpu/model.layers.2.mlp.dense_4h_to_h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:029775363050d9e4c232301274a8085474ca7992a42cabc1112ff46938c5a178 -size 16384 diff --git a/converter/testdata/1-gpu/model.layers.2.mlp.dense_h_to_4h.bias.0.bin b/converter/testdata/1-gpu/model.layers.2.mlp.dense_h_to_4h.bias.0.bin deleted file mode 100644 index c30ec76..0000000 --- a/converter/testdata/1-gpu/model.layers.2.mlp.dense_h_to_4h.bias.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:076a27c79e5ace2a3d47f9dd2e83e4ff6ea8872b3c2218f66c92b89b55f36560 -size 512 diff --git a/converter/testdata/1-gpu/model.layers.2.mlp.dense_h_to_4h.weight.0.bin b/converter/testdata/1-gpu/model.layers.2.mlp.dense_h_to_4h.weight.0.bin deleted file mode 100644 index 63df77d..0000000 --- a/converter/testdata/1-gpu/model.layers.2.mlp.dense_h_to_4h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:95beee5255e9d3d5d255ebc6d3f18465e1ab005b7e33e3d0d6495eb7a6178eed -size 16384 diff --git a/converter/testdata/1-gpu/model.layers.3.attention.dense.weight.0.bin b/converter/testdata/1-gpu/model.layers.3.attention.dense.weight.0.bin deleted file mode 100644 index dfc1df1..0000000 --- a/converter/testdata/1-gpu/model.layers.3.attention.dense.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:7af1303e224fc1185c5d936ca8fb3ab336e46f1bd09c94f1e749936d0b023713 -size 4096 diff --git a/converter/testdata/1-gpu/model.layers.3.attention.query_key_value.weight.0.bin b/converter/testdata/1-gpu/model.layers.3.attention.query_key_value.weight.0.bin deleted file mode 100644 index 2603376..0000000 --- a/converter/testdata/1-gpu/model.layers.3.attention.query_key_value.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:55961585790611913b727f6043af3589274c4b2350ea68bcacf25009e750cc37 -size 12288 diff --git a/converter/testdata/1-gpu/model.layers.3.input_layernorm.bias.bin b/converter/testdata/1-gpu/model.layers.3.input_layernorm.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.3.input_layernorm.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.3.input_layernorm.weight.bin b/converter/testdata/1-gpu/model.layers.3.input_layernorm.weight.bin deleted file mode 100644 index d0c1df9..0000000 --- a/converter/testdata/1-gpu/model.layers.3.input_layernorm.weight.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:b638277a8690e175a9137feff1e43c067f9faf4e2f600caf468fb05b0403b717 -size 128 diff --git a/converter/testdata/1-gpu/model.layers.3.mlp.dense_4h_to_h.bias.bin b/converter/testdata/1-gpu/model.layers.3.mlp.dense_4h_to_h.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.3.mlp.dense_4h_to_h.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.3.mlp.dense_4h_to_h.weight.0.bin b/converter/testdata/1-gpu/model.layers.3.mlp.dense_4h_to_h.weight.0.bin deleted file mode 100644 index 01f32ad..0000000 --- a/converter/testdata/1-gpu/model.layers.3.mlp.dense_4h_to_h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:9f37f3e28a6e716d08d29d02126854bd6e6248763b9a30af2f2e1dcf7b8fd9a5 -size 16384 diff --git a/converter/testdata/1-gpu/model.layers.3.mlp.dense_h_to_4h.bias.0.bin b/converter/testdata/1-gpu/model.layers.3.mlp.dense_h_to_4h.bias.0.bin deleted file mode 100644 index c30ec76..0000000 --- a/converter/testdata/1-gpu/model.layers.3.mlp.dense_h_to_4h.bias.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:076a27c79e5ace2a3d47f9dd2e83e4ff6ea8872b3c2218f66c92b89b55f36560 -size 512 diff --git a/converter/testdata/1-gpu/model.layers.3.mlp.dense_h_to_4h.weight.0.bin b/converter/testdata/1-gpu/model.layers.3.mlp.dense_h_to_4h.weight.0.bin deleted file mode 100644 index 10fcdeb..0000000 --- a/converter/testdata/1-gpu/model.layers.3.mlp.dense_h_to_4h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:d903dfb50cce4eae673a2d96a7814353ac1436c0c4d65490bb3ed7d8f8cfc586 -size 16384 diff --git a/converter/testdata/1-gpu/model.layers.4.attention.dense.weight.0.bin b/converter/testdata/1-gpu/model.layers.4.attention.dense.weight.0.bin deleted file mode 100644 index 2d07ce1..0000000 --- a/converter/testdata/1-gpu/model.layers.4.attention.dense.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:cb76390eda7af1d23d5b162446b4c5da0b791ec337d20d9a643b9716377ae514 -size 4096 diff --git a/converter/testdata/1-gpu/model.layers.4.attention.query_key_value.weight.0.bin b/converter/testdata/1-gpu/model.layers.4.attention.query_key_value.weight.0.bin deleted file mode 100644 index b0008a0..0000000 --- a/converter/testdata/1-gpu/model.layers.4.attention.query_key_value.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:8f54ff65f519de2db6343ec257da92091162381082fbb5dba04df50d8e12a158 -size 12288 diff --git a/converter/testdata/1-gpu/model.layers.4.input_layernorm.bias.bin b/converter/testdata/1-gpu/model.layers.4.input_layernorm.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.4.input_layernorm.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.4.input_layernorm.weight.bin b/converter/testdata/1-gpu/model.layers.4.input_layernorm.weight.bin deleted file mode 100644 index d0c1df9..0000000 --- a/converter/testdata/1-gpu/model.layers.4.input_layernorm.weight.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:b638277a8690e175a9137feff1e43c067f9faf4e2f600caf468fb05b0403b717 -size 128 diff --git a/converter/testdata/1-gpu/model.layers.4.mlp.dense_4h_to_h.bias.bin b/converter/testdata/1-gpu/model.layers.4.mlp.dense_4h_to_h.bias.bin deleted file mode 100644 index 67415c7..0000000 --- a/converter/testdata/1-gpu/model.layers.4.mlp.dense_4h_to_h.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:38723a2e5e8a17aa7950dc008209944e898f69a7bd10a23c839d341e935fd5ca -size 128 diff --git a/converter/testdata/1-gpu/model.layers.4.mlp.dense_4h_to_h.weight.0.bin b/converter/testdata/1-gpu/model.layers.4.mlp.dense_4h_to_h.weight.0.bin deleted file mode 100644 index 41b7c12..0000000 --- a/converter/testdata/1-gpu/model.layers.4.mlp.dense_4h_to_h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:67e7225a2de28f7e21a72b7ea1cab517071744c53d13f4c6964cbc27f4d261e4 -size 16384 diff --git a/converter/testdata/1-gpu/model.layers.4.mlp.dense_h_to_4h.bias.0.bin b/converter/testdata/1-gpu/model.layers.4.mlp.dense_h_to_4h.bias.0.bin deleted file mode 100644 index c30ec76..0000000 --- a/converter/testdata/1-gpu/model.layers.4.mlp.dense_h_to_4h.bias.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:076a27c79e5ace2a3d47f9dd2e83e4ff6ea8872b3c2218f66c92b89b55f36560 -size 512 diff --git a/converter/testdata/1-gpu/model.layers.4.mlp.dense_h_to_4h.weight.0.bin b/converter/testdata/1-gpu/model.layers.4.mlp.dense_h_to_4h.weight.0.bin deleted file mode 100644 index 975aa56..0000000 --- a/converter/testdata/1-gpu/model.layers.4.mlp.dense_h_to_4h.weight.0.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:cb7582613a3b40b73ac166af2e6b9773511cf6fba7bec12f340119f588c8ea48 -size 16384 diff --git a/converter/testdata/1-gpu/model.lm_head.bias.bin b/converter/testdata/1-gpu/model.lm_head.bias.bin deleted file mode 100644 index fb155d8..0000000 --- a/converter/testdata/1-gpu/model.lm_head.bias.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:fc19b1997119425765295aeab72d76faa6927d4f83985d328c26f20468d6cc76 -size 4000 diff --git a/converter/testdata/1-gpu/model.lm_head.weight.bin b/converter/testdata/1-gpu/model.lm_head.weight.bin deleted file mode 100644 index 530ff0d..0000000 --- a/converter/testdata/1-gpu/model.lm_head.weight.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:884fae8e52776e900d3800048825503048c95ab2850c38c1a595c6da962d3286 -size 128000 diff --git a/converter/testdata/1-gpu/model.wte.bin b/converter/testdata/1-gpu/model.wte.bin deleted file mode 100644 index c815f00..0000000 --- a/converter/testdata/1-gpu/model.wte.bin +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:d0d0fec061d1987d0f19acbd319c3471037691cebecd5aa2b00cedcdc2c6177c -size 128000