init commit

README.md

@@ -1,3 +1,135 @@
 ---
+language: ja
+thumbnail: https://github.com/rinnakk/japanese-pretrained-models/blob/master/rinna.png
+tags:
+- ja
+- gpt_neox
+- text-generation
+- lm
+- nlp
 license: mit
+datasets:
+- Anthropic/hh-rlhf
+- stanfordnlp/SHP
+inference: false
 ---
+
+# japanese-gpt-neox-3.6b-instruction-sft
+
+![rinna-icon](./rinna.png)
+
+This repository provides a Japanese GPT-NeoX model of 3.6 billion parameters. The model is based on [`rinna/japanese-gpt-neox-3.6b`](https://huggingface.co/rinna/japanese-gpt-neox-3.6b) and has been finetuned to serve as a instruction-following conversational agent.
+
+A special format has been adopted to construct inputs.
+* An input prompt is formatted as a conversation between `ユーザー` and `システム`.
+* Each input utterance consists of (1) its speaker (`"ユーザー"` or `"システム"`), (2) a colon (`":"`), (3) a whitespace (`" "`), and (4) utterance text (e.g. `"世界で一番高い山は？"`).
+* The input prompt should be ended with `"システム: "` to acknowledge the model to generate a response.
+* Since the model's tokenizer does not recognize `"\n"`, a special newline symbol `"<NL>"` is used instead.
+* All the newlines in input and output utterances should be replaced with `"<NL>"`.
+* All the utterances in the input prompt should be separated by `"<NL>"`.
+
+Following is an example to construct an input from a conversation.
+~~~python
+prompt = [
+    {
+        "speaker": "ユーザー",
+        "text": "日本のおすすめの観光地を教えてください。"
+    },
+    {
+        "speaker": "システム",
+        "text": "どの地域の観光地が知りたいですか？"
+    },
+    {
+        "speaker": "ユーザー",
+        "text": "渋谷の観光地を教えてください。"
+    }
+]
+prompt = [
+    f"{uttr['speaker']}: {uttr['text']}"
+    for uttr in prompt
+]
+prompt = "<NL>".join(prompt)
+prompt = (
+    prompt
+    + "<NL>"
+    + "システム: "
+)
+print(prompt)
+# "ユーザー: 日本のおすすめの観光地を教えてください。<NL>システム: どの地域の観光地が知りたいですか？<NL>ユーザー: 渋谷の観光地を教えてください。<NL>システム: "
+~~~
+
+# How to use the model
+
+~~~~python
+import torch
+from transformers import AutoTokenizer, AutoModelForCausalLM
+
+tokenizer = AutoTokenizer.from_pretrained(".", use_fast=False)
+model = AutoModelForCausalLM.from_pretrained(".")
+
+if torch.cuda.is_available():
+    model = model.to("cuda")
+
+token_ids = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt")
+
+with torch.no_grad():
+    output_ids = model.generate(
+        token_ids.to(model.device),
+        do_sample=True,
+        max_new_tokens=128,
+        temperature=0.7,
+        pad_token_id=tokenizer.pad_token_id,
+        bos_token_id=tokenizer.bos_token_id,
+        eos_token_id=tokenizer.eos_token_id
+    )
+
+output = tokenizer.decode(output_ids.tolist()[0][token_ids.size(1):])
+output = output.replace("<NL>", "\n")
+print(output)
+"""分かりました。いくつかのおすすめを紹介します。
+1. ハチ公像です。ハチ公像は、日本の観光スポットの1つとして人気があります。
+2. スクランブル交差点です。多くの人々が行き交う大きな交差点で、観光客に人気のスポットです。
+3. 109です。109は、ショッピングやエンターテイメント施設です。
+4. 道玄坂です。道玄坂は、日本の商業地区である坂道です。</s>"""
+~~~~
+
+# Model architecture
+A 36-layer, 2816-hidden-size transformer-based language model.
+
+# Finetuning
+The finetuning data is the subset of the following datasets and has been translated into Japanese.
+* [Anthropic HH RLHF data](https://huggingface.co/datasets/Anthropic/hh-rlhf)
+* [FLAN Instruction Tuning data](https://github.com/google-research/FLAN)
+* [Stanford Human Preferences Dataset](https://huggingface.co/datasets/stanfordnlp/SHP)
+
+The data will **not** be released.
+
+# Tokenization
+The model uses a [sentencepiece](https://github.com/google/sentencepiece)-based tokenizer.
+* The tokenizer has a vocabulary size of 32,000.
+* It uses sentencepiece's byte fallback feature to decompose unknown text pieces into UTF-8 byte pieces and to avoid producing `<UNK>` tokens.
+* sentencepiece's `--add_dummy_prefix` option was turned off so that a leading whitespace will not be prepended automatically.
+    ~~~
+    print(tokenizer.tokenize("吾輩は猫である"))
+    # ['吾', '輩', 'は', '猫', 'である']
+    # instead of ['▁', '吾', '輩', 'は', '猫', 'である'] as in rinna/japanese-gpt-1b
+    ~~~
+* sentencepiece's `--remove_extra_whitespaces` option was turned off so that leading, trailing, and duplicate whitespaces are reserved.
+    ~~~
+    print(tokenizer.tokenize("  吾輩は  猫である   "))
+    # ['▁', '▁', '吾', '輩', 'は', '▁', '▁', '猫', 'である', '▁', '▁', '▁']
+    # instead of ['▁', '吾', '輩', 'は', '▁猫', 'である'] as in rinna/japanese-gpt-1b
+    ~~~
+* Don't forget to set `use_fast=False` to make the above features function correctly.
+    ~~~
+    good_tokenizer = AutoTokenizer.from_pretrained("rinna/japanese-gpt-neox-3.6b", use_fast=False)
+    bad_tokenizer = AutoTokenizer.from_pretrained("rinna/japanese-gpt-neox-3.6b")
+
+    print(good_tokenizer.decode(good_tokenizer.encode("გამარჯობა  吾輩は  猫である   ")))
+    # 'გამარჯობა  吾輩は  猫である   </s>'
+    print(bad_tokenizer.decode(bad_tokenizer.encode("გამარჯობა  吾輩は  猫である   ")))
+    # 'გამარ[UNK]ობა 吾輩は 猫である </s>'
+    ~~~
+
+# Licenese
+[The MIT license](https://opensource.org/licenses/MIT)

config.json

@@ -0,0 +1,23 @@
+{
+  "architectures": [
+    "GPTNeoXForCausalLM"
+  ],
+  "bos_token_id": 2,
+  "eos_token_id": 3,
+  "hidden_act": "gelu",
+  "hidden_size": 2816,
+  "initializer_range": 0.02,
+  "intermediate_size": 11264,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 2048,
+  "model_type": "gpt_neox",
+  "num_attention_heads": 22,
+  "num_hidden_layers": 36,
+  "rotary_emb_base": 10000,
+  "rotary_pct": 1.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "use_cache": true,
+  "use_parallel_residual": false,
+  "vocab_size": 32000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c6124c628f8ecc29be1b6ee0625670062340f5b99cfe543ccf049fa90e6207b
+size 7365670537

spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d78ab344146700112cd41628ac7ce54b79c0868fe0c7c201750d8237b54dbb4
+size 786216

tokenizer_config.json

@@ -0,0 +1 @@
+{"eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]", "extra_ids": 0, "additional_special_tokens": [], "sp_model_kwargs": {}, "bos_token": "<s>", "cls_token": "[CLS]", "sep_token": "[SEP]", "mask_token": "[MASK]", "do_lower_case": false, "tokenizer_class": "T5Tokenizer"}