Upload tiny models for GPTSanJapaneseForConditionalGeneration

config.json

@@ -0,0 +1,33 @@
+{
+  "architectures": [
+    "GPTSanJapaneseForConditionalGeneration"
+  ],
+  "d_ext": 80,
+  "d_ff": 32,
+  "d_model": 32,
+  "d_spout": 33,
+  "dropout_rate": 0.0,
+  "eos_token_id": 35999,
+  "expert_capacity": 100,
+  "initializer_factor": 0.002,
+  "layer_norm_epsilon": 1e-06,
+  "max_position_embeddings": 1280,
+  "model_type": "gptsan-japanese",
+  "num_contexts": 7,
+  "num_experts": 2,
+  "num_ext_layers": 2,
+  "num_heads": 4,
+  "num_layers": 5,
+  "num_switch_layers": 3,
+  "output_router_logits": false,
+  "pad_token_id": 35999,
+  "router_bias": false,
+  "router_dtype": "float32",
+  "router_ignore_padding_tokens": false,
+  "router_jitter_noise": 0.0,
+  "separator_token_id": 35998,
+  "torch_dtype": "float32",
+  "transformers_version": "4.28.0.dev0",
+  "use_cache": true,
+  "vocab_size": 35998
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "eos_token_id": 35999,
+  "pad_token_id": 35999,
+  "transformers_version": "4.28.0.dev0"
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd2c241e6adb9453f0a843977b61885c0d73ef63a79c6c8817c0faf115ef12f2
+size 5375076

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "bos_token": "<|startoftext|>",
+  "eos_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>",
+  "sep_token": "<|segmenter|>",
+  "unk_token": "<|endoftext|>"
+}

tokenizer_config.json

@@ -0,0 +1,12 @@
+{
+  "bos_token": "<|startoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "do_clean_text": false,
+  "eos_token": "<|endoftext|>",
+  "model_max_length": 1280,
+  "pad_token": "<|endoftext|>",
+  "sep_token": "<|segmenter|>",
+  "special_tokens_map_file": null,
+  "tokenizer_class": "GPTSanJapaneseTokenizer",
+  "unk_token": "<|endoftext|>"
+}