Hebrew book summary generator

README.md

@@ -1,12 +1,12 @@
 ---
-title: HebrewBookSummaries
-emoji: 🐠
+title: תקצירי ספרים
+emoji: 📚
 colorFrom: green
-colorTo: pink
+colorTo: red
 sdk: gradio
-sdk_version: 3.29.0
+sdk_version: 3.28.1
 app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,61 @@
+import gradio as gr
+import torch
+from transformers import pipeline, set_seed, AutoTokenizer, AutoModelForCausalLM
+import random
+
+title = "מחולל תקצירי ספרים מבוסס ג׳פיטי-נאו פצפון"
+article = "מודל השפה אומן על ידי <a href=\"https://linktr.ee/Norod78\">דורון אדלר</a>"
+description = "<p>Fine tuned <a href=\"https://huggingface.co/Norod78/hebrew-gpt_neo-tiny\">Norod78/hebrew-gpt_neo-tiny</a> upon a book summary dataset<p>"
+
+examples = [
+    ['אם מתחשק לכם לפעמים'],
+    ["מחוללי הטקסט / "],
+    ['האדם האחרון עלי אדמות ישב לבד בחדרו כשלפתע נשמע דפיקה'],
+]
+
+model_id = "./hebrew-gpt_neo-tiny-HebrewBookSummaries"
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+n_gpu = 0 if torch.cuda.is_available()==False else torch.cuda.device_count()
+tokenizer = AutoTokenizer.from_pretrained(model_id, bos_token='<|startoftext|>', eos_token='<|endoftext|>', pad_token='<|pad|>', unknown_token = '<|unknown|>')
+model = AutoModelForCausalLM.from_pretrained(model_id).to(device)
+model.resize_token_embeddings(len(tokenizer))
+text_generator = pipeline('text-generation', model=model, tokenizer=tokenizer)
+
+max_length = 96
+top_k = 40
+top_p = 0.92
+temperature = 0.8
+max_seed = (2**32)-1
+global_seed = random.randint(0, max_seed)
+
+def text_generation(input_text = ''):
+    global global_seed
+    global_seed = global_seed + 1
+    if global_seed >= max_seed:
+        global_seed = 0
+    if input_text == None or len(input_text) == 0:
+        input_text = "<|startoftext|>"
+    set_seed(global_seed)
+    generated_text = text_generator(input_text,
+    max_length=max_length,
+    top_k=top_k, 
+    top_p=top_p,    
+    temperature=temperature,
+    do_sample=True,
+    repetition_penalty=2.0,
+    num_return_sequences=1)
+    parsed_text = generated_text[0]["generated_text"].replace("<|startoftext|>", "").replace("\r","\n").replace("\n\n", "\n").replace("\t", " ").replace("<|pad|>", " * ").replace("\"\"", "\"").strip()
+    print("parsed_text = \"" + parsed_text + "\" (seed = " + str(global_seed) + ")")
+    return parsed_text
+gr.Interface(
+    text_generation,    
+    inputs=gr.Textbox(lines=1, label=".הזינו פה את מילות הפתיחה של הטקסט, בחרו את אחת מן הדוגמאות המוכנות או השאירו ריק. מה שבא לכם. בכל לחיצה על סאבמיט, יווצר טקסט אחר", elem_id="input_text"),
+    outputs=gr.Textbox(type="text", label="פה מופיע הטקסט שהמחולל יוצר", elem_id="output_text"),
+    css="#output_text{direction: rtl} #input_text{direction: rtl}",
+    title=title,
+    description=description,
+    article=article,
+    examples=examples,
+    allow_flagging='never',
+).launch()

hebrew-gpt_neo-tiny-HebrewBookSummaries/added_tokens.json

@@ -0,0 +1,6 @@
+{
+  "<|endoftext|>": 50257,
+  "<|pad|>": 50260,
+  "<|startoftext|>": 50258,
+  "<|unknown|>": 50259
+}

hebrew-gpt_neo-tiny-HebrewBookSummaries/config.json

@@ -0,0 +1,49 @@
+{
+  "_name_or_path": "Norod78/hebrew-gpt_neo-tiny",
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPTNeoForCausalLM"
+  ],
+  "attention_dropout": 0.1,
+  "attention_layers": [
+    "global",
+    "global",
+    "global",
+    "global",
+    "global",
+    "global"
+  ],
+  "attention_types": [
+    [
+      [
+        "global"
+      ],
+      6
+    ]
+  ],
+  "bos_token_id": 50256,
+  "classifier_dropout": 0.1,
+  "embed_dropout": 0.1,
+  "eos_token_id": 50256,
+  "gradient_checkpointing": false,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": null,
+  "layer_norm_epsilon": 1e-05,
+  "max_position_embeddings": 1024,
+  "model_type": "gpt_neo",
+  "num_heads": 12,
+  "num_layers": 6,
+  "pad_token_id": 50256,
+  "resid_dropout": 0.1,
+  "summary_activation": null,
+  "summary_first_dropout": 0.1,
+  "summary_proj_to_labels": true,
+  "summary_type": "cls_index",
+  "summary_use_proj": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.29.1",
+  "use_cache": true,
+  "vocab_size": 50261,
+  "window_size": 256
+}

hebrew-gpt_neo-tiny-HebrewBookSummaries/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 50256,
+  "eos_token_id": 50256,
+  "pad_token_id": 50256,
+  "transformers_version": "4.29.1"
+}

hebrew-gpt_neo-tiny-HebrewBookSummaries/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1710bf6e58073a6715acc83a80391274813046322f9812f0ef56844a2ef5881
+size 333929547

hebrew-gpt_neo-tiny-HebrewBookSummaries/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|pad|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": "<unk>"
+}

hebrew-gpt_neo-tiny-HebrewBookSummaries/tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "clean_up_tokenization_spaces": true,
+  "do_lower_case": false,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "errors": "replace",
+  "full_tokenizer_file": null,
+  "max_len": 1024,
+  "model_max_length": 1024,
+  "pad_token": {
+    "__type": "AddedToken",
+    "content": "<|pad|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unknown_token": "<|unknown|>"
+}

requirements.txt

@@ -0,0 +1,5 @@
+gradio
+torch
+transformers
+tokenizers
+