Revert "Making the vocab_size match the tokenizer vocab_size."
This reverts commit 4be554b688.
This commit is contained in:
Nicolas Patry
parent
4be554b688
commit
541d169ca9
574
config.json
574
config.json
|
|
@ -1,294 +1,294 @@
|
||||||
{
|
{
|
||||||
|
"initializer_factor": 1.0,
|
||||||
|
"logit_scale_init_value": 2.6592,
|
||||||
|
"model_type": "clip",
|
||||||
|
"projection_dim": 64,
|
||||||
|
"text_config": {
|
||||||
|
"_name_or_path": "",
|
||||||
|
"add_cross_attention": false,
|
||||||
|
"architectures": null,
|
||||||
|
"attention_dropout": 0.1,
|
||||||
|
"bad_words_ids": null,
|
||||||
|
"bos_token_id": 0,
|
||||||
|
"chunk_size_feed_forward": 0,
|
||||||
|
"decoder_start_token_id": null,
|
||||||
|
"diversity_penalty": 0.0,
|
||||||
|
"do_sample": false,
|
||||||
|
"dropout": 0.1,
|
||||||
|
"early_stopping": false,
|
||||||
|
"encoder_no_repeat_ngram_size": 0,
|
||||||
|
"eos_token_id": 2,
|
||||||
|
"finetuning_task": null,
|
||||||
|
"forced_bos_token_id": null,
|
||||||
|
"forced_eos_token_id": null,
|
||||||
|
"gradient_checkpointing": false,
|
||||||
|
"hidden_act": "quick_gelu",
|
||||||
|
"hidden_size": 32,
|
||||||
|
"id2label": {
|
||||||
|
"0": "LABEL_0",
|
||||||
|
"1": "LABEL_1"
|
||||||
|
},
|
||||||
"initializer_factor": 1.0,
|
"initializer_factor": 1.0,
|
||||||
"logit_scale_init_value": 2.6592,
|
"initializer_range": 0.02,
|
||||||
"model_type": "clip",
|
"intermediate_size": 37,
|
||||||
"projection_dim": 64,
|
"is_decoder": false,
|
||||||
"text_config": {
|
"is_encoder_decoder": false,
|
||||||
"_name_or_path": "",
|
"label2id": {
|
||||||
"add_cross_attention": false,
|
"LABEL_0": 0,
|
||||||
"architectures": null,
|
"LABEL_1": 1
|
||||||
"attention_dropout": 0.1,
|
|
||||||
"bad_words_ids": null,
|
|
||||||
"bos_token_id": 0,
|
|
||||||
"chunk_size_feed_forward": 0,
|
|
||||||
"decoder_start_token_id": null,
|
|
||||||
"diversity_penalty": 0.0,
|
|
||||||
"do_sample": false,
|
|
||||||
"dropout": 0.1,
|
|
||||||
"early_stopping": false,
|
|
||||||
"encoder_no_repeat_ngram_size": 0,
|
|
||||||
"eos_token_id": 2,
|
|
||||||
"finetuning_task": null,
|
|
||||||
"forced_bos_token_id": null,
|
|
||||||
"forced_eos_token_id": null,
|
|
||||||
"gradient_checkpointing": false,
|
|
||||||
"hidden_act": "quick_gelu",
|
|
||||||
"hidden_size": 32,
|
|
||||||
"id2label": {
|
|
||||||
"0": "LABEL_0",
|
|
||||||
"1": "LABEL_1"
|
|
||||||
},
|
|
||||||
"initializer_factor": 1.0,
|
|
||||||
"initializer_range": 0.02,
|
|
||||||
"intermediate_size": 37,
|
|
||||||
"is_decoder": false,
|
|
||||||
"is_encoder_decoder": false,
|
|
||||||
"label2id": {
|
|
||||||
"LABEL_0": 0,
|
|
||||||
"LABEL_1": 1
|
|
||||||
},
|
|
||||||
"layer_norm_eps": 1e-5,
|
|
||||||
"length_penalty": 1.0,
|
|
||||||
"max_length": 20,
|
|
||||||
"max_position_embeddings": 512,
|
|
||||||
"min_length": 0,
|
|
||||||
"model_type": "clip_text_model",
|
|
||||||
"no_repeat_ngram_size": 0,
|
|
||||||
"num_attention_heads": 4,
|
|
||||||
"num_beam_groups": 1,
|
|
||||||
"num_beams": 1,
|
|
||||||
"num_hidden_layers": 5,
|
|
||||||
"num_return_sequences": 1,
|
|
||||||
"output_attentions": false,
|
|
||||||
"output_hidden_states": false,
|
|
||||||
"output_scores": false,
|
|
||||||
"pad_token_id": 1,
|
|
||||||
"prefix": null,
|
|
||||||
"problem_type": null,
|
|
||||||
"pruned_heads": {},
|
|
||||||
"remove_invalid_values": false,
|
|
||||||
"repetition_penalty": 1.0,
|
|
||||||
"return_dict": true,
|
|
||||||
"return_dict_in_generate": false,
|
|
||||||
"sep_token_id": null,
|
|
||||||
"task_specific_params": null,
|
|
||||||
"temperature": 1.0,
|
|
||||||
"tie_encoder_decoder": false,
|
|
||||||
"tie_word_embeddings": true,
|
|
||||||
"tokenizer_class": null,
|
|
||||||
"top_k": 50,
|
|
||||||
"top_p": 1.0,
|
|
||||||
"torch_dtype": null,
|
|
||||||
"torchscript": false,
|
|
||||||
"transformers_version": "4.11.0.dev0",
|
|
||||||
"use_bfloat16": false,
|
|
||||||
"vocab_size": 1000
|
|
||||||
},
|
},
|
||||||
"text_config_dict": {
|
"layer_norm_eps": 1e-05,
|
||||||
"_name_or_path": "",
|
"length_penalty": 1.0,
|
||||||
"add_cross_attention": false,
|
"max_length": 20,
|
||||||
"architectures": null,
|
"max_position_embeddings": 512,
|
||||||
"attention_dropout": 0.1,
|
"min_length": 0,
|
||||||
"bad_words_ids": null,
|
"model_type": "clip_text_model",
|
||||||
"bos_token_id": 0,
|
"no_repeat_ngram_size": 0,
|
||||||
"chunk_size_feed_forward": 0,
|
"num_attention_heads": 4,
|
||||||
"decoder_start_token_id": null,
|
"num_beam_groups": 1,
|
||||||
"diversity_penalty": 0.0,
|
"num_beams": 1,
|
||||||
"do_sample": false,
|
"num_hidden_layers": 5,
|
||||||
"dropout": 0.1,
|
"num_return_sequences": 1,
|
||||||
"early_stopping": false,
|
"output_attentions": false,
|
||||||
"encoder_no_repeat_ngram_size": 0,
|
"output_hidden_states": false,
|
||||||
"eos_token_id": 2,
|
"output_scores": false,
|
||||||
"finetuning_task": null,
|
"pad_token_id": 1,
|
||||||
"forced_bos_token_id": null,
|
"prefix": null,
|
||||||
"forced_eos_token_id": null,
|
"problem_type": null,
|
||||||
"gradient_checkpointing": false,
|
"pruned_heads": {},
|
||||||
"hidden_act": "quick_gelu",
|
"remove_invalid_values": false,
|
||||||
"hidden_size": 32,
|
"repetition_penalty": 1.0,
|
||||||
"id2label": {
|
"return_dict": true,
|
||||||
"0": "LABEL_0",
|
"return_dict_in_generate": false,
|
||||||
"1": "LABEL_1"
|
"sep_token_id": null,
|
||||||
},
|
"task_specific_params": null,
|
||||||
"initializer_factor": 1.0,
|
"temperature": 1.0,
|
||||||
"initializer_range": 0.02,
|
"tie_encoder_decoder": false,
|
||||||
"intermediate_size": 37,
|
"tie_word_embeddings": true,
|
||||||
"is_decoder": false,
|
"tokenizer_class": null,
|
||||||
"is_encoder_decoder": false,
|
"top_k": 50,
|
||||||
"label2id": {
|
"top_p": 1.0,
|
||||||
"LABEL_0": 0,
|
"torch_dtype": null,
|
||||||
"LABEL_1": 1
|
"torchscript": false,
|
||||||
},
|
"transformers_version": "4.11.0.dev0",
|
||||||
"layer_norm_eps": 1e-5,
|
"use_bfloat16": false,
|
||||||
"length_penalty": 1.0,
|
"vocab_size": 99
|
||||||
"max_length": 20,
|
},
|
||||||
"max_position_embeddings": 512,
|
"text_config_dict": {
|
||||||
"min_length": 0,
|
"_name_or_path": "",
|
||||||
"model_type": "clip_text_model",
|
"add_cross_attention": false,
|
||||||
"no_repeat_ngram_size": 0,
|
"architectures": null,
|
||||||
"num_attention_heads": 4,
|
"attention_dropout": 0.1,
|
||||||
"num_beam_groups": 1,
|
"bad_words_ids": null,
|
||||||
"num_beams": 1,
|
"bos_token_id": 0,
|
||||||
"num_hidden_layers": 5,
|
"chunk_size_feed_forward": 0,
|
||||||
"num_return_sequences": 1,
|
"decoder_start_token_id": null,
|
||||||
"output_attentions": false,
|
"diversity_penalty": 0.0,
|
||||||
"output_hidden_states": false,
|
"do_sample": false,
|
||||||
"output_scores": false,
|
"dropout": 0.1,
|
||||||
"pad_token_id": 1,
|
"early_stopping": false,
|
||||||
"prefix": null,
|
"encoder_no_repeat_ngram_size": 0,
|
||||||
"problem_type": null,
|
"eos_token_id": 2,
|
||||||
"pruned_heads": {},
|
"finetuning_task": null,
|
||||||
"remove_invalid_values": false,
|
"forced_bos_token_id": null,
|
||||||
"repetition_penalty": 1.0,
|
"forced_eos_token_id": null,
|
||||||
"return_dict": true,
|
"gradient_checkpointing": false,
|
||||||
"return_dict_in_generate": false,
|
"hidden_act": "quick_gelu",
|
||||||
"sep_token_id": null,
|
"hidden_size": 32,
|
||||||
"task_specific_params": null,
|
"id2label": {
|
||||||
"temperature": 1.0,
|
"0": "LABEL_0",
|
||||||
"tie_encoder_decoder": false,
|
"1": "LABEL_1"
|
||||||
"tie_word_embeddings": true,
|
|
||||||
"tokenizer_class": null,
|
|
||||||
"top_k": 50,
|
|
||||||
"top_p": 1.0,
|
|
||||||
"torch_dtype": null,
|
|
||||||
"torchscript": false,
|
|
||||||
"transformers_version": "4.11.0.dev0",
|
|
||||||
"use_bfloat16": false,
|
|
||||||
"vocab_size": 1000
|
|
||||||
},
|
},
|
||||||
"transformers_version": null,
|
"initializer_factor": 1.0,
|
||||||
"vision_config": {
|
"initializer_range": 0.02,
|
||||||
"_name_or_path": "",
|
"intermediate_size": 37,
|
||||||
"add_cross_attention": false,
|
"is_decoder": false,
|
||||||
"architectures": null,
|
"is_encoder_decoder": false,
|
||||||
"attention_dropout": 0.1,
|
"label2id": {
|
||||||
"bad_words_ids": null,
|
"LABEL_0": 0,
|
||||||
"bos_token_id": null,
|
"LABEL_1": 1
|
||||||
"chunk_size_feed_forward": 0,
|
|
||||||
"decoder_start_token_id": null,
|
|
||||||
"diversity_penalty": 0.0,
|
|
||||||
"do_sample": false,
|
|
||||||
"dropout": 0.1,
|
|
||||||
"early_stopping": false,
|
|
||||||
"encoder_no_repeat_ngram_size": 0,
|
|
||||||
"eos_token_id": null,
|
|
||||||
"finetuning_task": null,
|
|
||||||
"forced_bos_token_id": null,
|
|
||||||
"forced_eos_token_id": null,
|
|
||||||
"gradient_checkpointing": false,
|
|
||||||
"hidden_act": "quick_gelu",
|
|
||||||
"hidden_size": 32,
|
|
||||||
"id2label": {
|
|
||||||
"0": "LABEL_0",
|
|
||||||
"1": "LABEL_1"
|
|
||||||
},
|
|
||||||
"image_size": 30,
|
|
||||||
"initializer_factor": 1.0,
|
|
||||||
"initializer_range": 0.02,
|
|
||||||
"intermediate_size": 37,
|
|
||||||
"is_decoder": false,
|
|
||||||
"is_encoder_decoder": false,
|
|
||||||
"label2id": {
|
|
||||||
"LABEL_0": 0,
|
|
||||||
"LABEL_1": 1
|
|
||||||
},
|
|
||||||
"layer_norm_eps": 1e-5,
|
|
||||||
"length_penalty": 1.0,
|
|
||||||
"max_length": 20,
|
|
||||||
"min_length": 0,
|
|
||||||
"model_type": "clip_vision_model",
|
|
||||||
"no_repeat_ngram_size": 0,
|
|
||||||
"num_attention_heads": 4,
|
|
||||||
"num_beam_groups": 1,
|
|
||||||
"num_beams": 1,
|
|
||||||
"num_channels": 3,
|
|
||||||
"num_hidden_layers": 5,
|
|
||||||
"num_return_sequences": 1,
|
|
||||||
"output_attentions": false,
|
|
||||||
"output_hidden_states": false,
|
|
||||||
"output_scores": false,
|
|
||||||
"pad_token_id": null,
|
|
||||||
"patch_size": 2,
|
|
||||||
"prefix": null,
|
|
||||||
"problem_type": null,
|
|
||||||
"pruned_heads": {},
|
|
||||||
"remove_invalid_values": false,
|
|
||||||
"repetition_penalty": 1.0,
|
|
||||||
"return_dict": true,
|
|
||||||
"return_dict_in_generate": false,
|
|
||||||
"sep_token_id": null,
|
|
||||||
"task_specific_params": null,
|
|
||||||
"temperature": 1.0,
|
|
||||||
"tie_encoder_decoder": false,
|
|
||||||
"tie_word_embeddings": true,
|
|
||||||
"tokenizer_class": null,
|
|
||||||
"top_k": 50,
|
|
||||||
"top_p": 1.0,
|
|
||||||
"torch_dtype": null,
|
|
||||||
"torchscript": false,
|
|
||||||
"transformers_version": "4.11.0.dev0",
|
|
||||||
"use_bfloat16": false
|
|
||||||
},
|
},
|
||||||
"vision_config_dict": {
|
"layer_norm_eps": 1e-05,
|
||||||
"_name_or_path": "",
|
"length_penalty": 1.0,
|
||||||
"add_cross_attention": false,
|
"max_length": 20,
|
||||||
"architectures": null,
|
"max_position_embeddings": 512,
|
||||||
"attention_dropout": 0.1,
|
"min_length": 0,
|
||||||
"bad_words_ids": null,
|
"model_type": "clip_text_model",
|
||||||
"bos_token_id": null,
|
"no_repeat_ngram_size": 0,
|
||||||
"chunk_size_feed_forward": 0,
|
"num_attention_heads": 4,
|
||||||
"decoder_start_token_id": null,
|
"num_beam_groups": 1,
|
||||||
"diversity_penalty": 0.0,
|
"num_beams": 1,
|
||||||
"do_sample": false,
|
"num_hidden_layers": 5,
|
||||||
"dropout": 0.1,
|
"num_return_sequences": 1,
|
||||||
"early_stopping": false,
|
"output_attentions": false,
|
||||||
"encoder_no_repeat_ngram_size": 0,
|
"output_hidden_states": false,
|
||||||
"eos_token_id": null,
|
"output_scores": false,
|
||||||
"finetuning_task": null,
|
"pad_token_id": 1,
|
||||||
"forced_bos_token_id": null,
|
"prefix": null,
|
||||||
"forced_eos_token_id": null,
|
"problem_type": null,
|
||||||
"gradient_checkpointing": false,
|
"pruned_heads": {},
|
||||||
"hidden_act": "quick_gelu",
|
"remove_invalid_values": false,
|
||||||
"hidden_size": 32,
|
"repetition_penalty": 1.0,
|
||||||
"id2label": {
|
"return_dict": true,
|
||||||
"0": "LABEL_0",
|
"return_dict_in_generate": false,
|
||||||
"1": "LABEL_1"
|
"sep_token_id": null,
|
||||||
},
|
"task_specific_params": null,
|
||||||
"image_size": 30,
|
"temperature": 1.0,
|
||||||
"initializer_factor": 1.0,
|
"tie_encoder_decoder": false,
|
||||||
"initializer_range": 0.02,
|
"tie_word_embeddings": true,
|
||||||
"intermediate_size": 37,
|
"tokenizer_class": null,
|
||||||
"is_decoder": false,
|
"top_k": 50,
|
||||||
"is_encoder_decoder": false,
|
"top_p": 1.0,
|
||||||
"label2id": {
|
"torch_dtype": null,
|
||||||
"LABEL_0": 0,
|
"torchscript": false,
|
||||||
"LABEL_1": 1
|
"transformers_version": "4.11.0.dev0",
|
||||||
},
|
"use_bfloat16": false,
|
||||||
"layer_norm_eps": 1e-5,
|
"vocab_size": 99
|
||||||
"length_penalty": 1.0,
|
},
|
||||||
"max_length": 20,
|
"transformers_version": null,
|
||||||
"min_length": 0,
|
"vision_config": {
|
||||||
"model_type": "clip_vision_model",
|
"_name_or_path": "",
|
||||||
"no_repeat_ngram_size": 0,
|
"add_cross_attention": false,
|
||||||
"num_attention_heads": 4,
|
"architectures": null,
|
||||||
"num_beam_groups": 1,
|
"attention_dropout": 0.1,
|
||||||
"num_beams": 1,
|
"bad_words_ids": null,
|
||||||
"num_channels": 3,
|
"bos_token_id": null,
|
||||||
"num_hidden_layers": 5,
|
"chunk_size_feed_forward": 0,
|
||||||
"num_return_sequences": 1,
|
"decoder_start_token_id": null,
|
||||||
"output_attentions": false,
|
"diversity_penalty": 0.0,
|
||||||
"output_hidden_states": false,
|
"do_sample": false,
|
||||||
"output_scores": false,
|
"dropout": 0.1,
|
||||||
"pad_token_id": null,
|
"early_stopping": false,
|
||||||
"patch_size": 2,
|
"encoder_no_repeat_ngram_size": 0,
|
||||||
"prefix": null,
|
"eos_token_id": null,
|
||||||
"problem_type": null,
|
"finetuning_task": null,
|
||||||
"pruned_heads": {},
|
"forced_bos_token_id": null,
|
||||||
"remove_invalid_values": false,
|
"forced_eos_token_id": null,
|
||||||
"repetition_penalty": 1.0,
|
"gradient_checkpointing": false,
|
||||||
"return_dict": true,
|
"hidden_act": "quick_gelu",
|
||||||
"return_dict_in_generate": false,
|
"hidden_size": 32,
|
||||||
"sep_token_id": null,
|
"id2label": {
|
||||||
"task_specific_params": null,
|
"0": "LABEL_0",
|
||||||
"temperature": 1.0,
|
"1": "LABEL_1"
|
||||||
"tie_encoder_decoder": false,
|
|
||||||
"tie_word_embeddings": true,
|
|
||||||
"tokenizer_class": null,
|
|
||||||
"top_k": 50,
|
|
||||||
"top_p": 1.0,
|
|
||||||
"torch_dtype": null,
|
|
||||||
"torchscript": false,
|
|
||||||
"transformers_version": "4.11.0.dev0",
|
|
||||||
"use_bfloat16": false
|
|
||||||
},
|
},
|
||||||
"vocab_size": 1000
|
"image_size": 30,
|
||||||
|
"initializer_factor": 1.0,
|
||||||
|
"initializer_range": 0.02,
|
||||||
|
"intermediate_size": 37,
|
||||||
|
"is_decoder": false,
|
||||||
|
"is_encoder_decoder": false,
|
||||||
|
"label2id": {
|
||||||
|
"LABEL_0": 0,
|
||||||
|
"LABEL_1": 1
|
||||||
|
},
|
||||||
|
"layer_norm_eps": 1e-05,
|
||||||
|
"length_penalty": 1.0,
|
||||||
|
"max_length": 20,
|
||||||
|
"min_length": 0,
|
||||||
|
"model_type": "clip_vision_model",
|
||||||
|
"no_repeat_ngram_size": 0,
|
||||||
|
"num_attention_heads": 4,
|
||||||
|
"num_beam_groups": 1,
|
||||||
|
"num_beams": 1,
|
||||||
|
"num_channels": 3,
|
||||||
|
"num_hidden_layers": 5,
|
||||||
|
"num_return_sequences": 1,
|
||||||
|
"output_attentions": false,
|
||||||
|
"output_hidden_states": false,
|
||||||
|
"output_scores": false,
|
||||||
|
"pad_token_id": null,
|
||||||
|
"patch_size": 2,
|
||||||
|
"prefix": null,
|
||||||
|
"problem_type": null,
|
||||||
|
"pruned_heads": {},
|
||||||
|
"remove_invalid_values": false,
|
||||||
|
"repetition_penalty": 1.0,
|
||||||
|
"return_dict": true,
|
||||||
|
"return_dict_in_generate": false,
|
||||||
|
"sep_token_id": null,
|
||||||
|
"task_specific_params": null,
|
||||||
|
"temperature": 1.0,
|
||||||
|
"tie_encoder_decoder": false,
|
||||||
|
"tie_word_embeddings": true,
|
||||||
|
"tokenizer_class": null,
|
||||||
|
"top_k": 50,
|
||||||
|
"top_p": 1.0,
|
||||||
|
"torch_dtype": null,
|
||||||
|
"torchscript": false,
|
||||||
|
"transformers_version": "4.11.0.dev0",
|
||||||
|
"use_bfloat16": false
|
||||||
|
},
|
||||||
|
"vision_config_dict": {
|
||||||
|
"_name_or_path": "",
|
||||||
|
"add_cross_attention": false,
|
||||||
|
"architectures": null,
|
||||||
|
"attention_dropout": 0.1,
|
||||||
|
"bad_words_ids": null,
|
||||||
|
"bos_token_id": null,
|
||||||
|
"chunk_size_feed_forward": 0,
|
||||||
|
"decoder_start_token_id": null,
|
||||||
|
"diversity_penalty": 0.0,
|
||||||
|
"do_sample": false,
|
||||||
|
"dropout": 0.1,
|
||||||
|
"early_stopping": false,
|
||||||
|
"encoder_no_repeat_ngram_size": 0,
|
||||||
|
"eos_token_id": null,
|
||||||
|
"finetuning_task": null,
|
||||||
|
"forced_bos_token_id": null,
|
||||||
|
"forced_eos_token_id": null,
|
||||||
|
"gradient_checkpointing": false,
|
||||||
|
"hidden_act": "quick_gelu",
|
||||||
|
"hidden_size": 32,
|
||||||
|
"id2label": {
|
||||||
|
"0": "LABEL_0",
|
||||||
|
"1": "LABEL_1"
|
||||||
|
},
|
||||||
|
"image_size": 30,
|
||||||
|
"initializer_factor": 1.0,
|
||||||
|
"initializer_range": 0.02,
|
||||||
|
"intermediate_size": 37,
|
||||||
|
"is_decoder": false,
|
||||||
|
"is_encoder_decoder": false,
|
||||||
|
"label2id": {
|
||||||
|
"LABEL_0": 0,
|
||||||
|
"LABEL_1": 1
|
||||||
|
},
|
||||||
|
"layer_norm_eps": 1e-05,
|
||||||
|
"length_penalty": 1.0,
|
||||||
|
"max_length": 20,
|
||||||
|
"min_length": 0,
|
||||||
|
"model_type": "clip_vision_model",
|
||||||
|
"no_repeat_ngram_size": 0,
|
||||||
|
"num_attention_heads": 4,
|
||||||
|
"num_beam_groups": 1,
|
||||||
|
"num_beams": 1,
|
||||||
|
"num_channels": 3,
|
||||||
|
"num_hidden_layers": 5,
|
||||||
|
"num_return_sequences": 1,
|
||||||
|
"output_attentions": false,
|
||||||
|
"output_hidden_states": false,
|
||||||
|
"output_scores": false,
|
||||||
|
"pad_token_id": null,
|
||||||
|
"patch_size": 2,
|
||||||
|
"prefix": null,
|
||||||
|
"problem_type": null,
|
||||||
|
"pruned_heads": {},
|
||||||
|
"remove_invalid_values": false,
|
||||||
|
"repetition_penalty": 1.0,
|
||||||
|
"return_dict": true,
|
||||||
|
"return_dict_in_generate": false,
|
||||||
|
"sep_token_id": null,
|
||||||
|
"task_specific_params": null,
|
||||||
|
"temperature": 1.0,
|
||||||
|
"tie_encoder_decoder": false,
|
||||||
|
"tie_word_embeddings": true,
|
||||||
|
"tokenizer_class": null,
|
||||||
|
"top_k": 50,
|
||||||
|
"top_p": 1.0,
|
||||||
|
"torch_dtype": null,
|
||||||
|
"torchscript": false,
|
||||||
|
"transformers_version": "4.11.0.dev0",
|
||||||
|
"use_bfloat16": false
|
||||||
|
},
|
||||||
|
"vocab_size": 1000
|
||||||
}
|
}
|
||||||
|
|
|
||||||
Loading…
Reference in New Issue