Upload files for streamlit demo

app.py

@@ -0,0 +1,41 @@
+import torch
+import streamlit as st
+from transformers import GPT2Tokenizer, GPT2LMHeadModel, GPT2TokenizerFast
+import numpy as np
+import jax
+import jax.numpy as jnp
+
+st.title("GPT2-korean")
+# st.markdown()
+
+seed = st.text_input("Seed", "안녕하세요")
+go = st.button("Generate")
+
+if go:
+    model = GPT2LMHeadModel.from_pretrained(".", pad_token_id=50256, from_flax=True)
+    tokenizer = GPT2TokenizerFast.from_pretrained(
+        ".", padding_side="left", pad_token="<|endoftext|>"
+    )
+    input_context = seed
+    input_ids = tokenizer(input_context, return_tensors="pt")
+    outputs = model.generate(
+        input_ids=input_ids["input_ids"],
+        max_length=50,
+        num_return_sequences=1,
+        num_beams=3,
+        no_repeat_ngram_size=3,
+        repetition_penalty=2.0,
+        do_sample=True,
+        bad_words_ids=[
+            [95],
+            [5470],
+            [504],
+            [528],
+            [504],
+            [919],
+            [65, 20374, 63],
+            [655],
+        ],
+    )
+    st.write(tokenizer.batch_decode(outputs, skip_special_tokens=True))
+

config.json

@@ -0,0 +1,37 @@
+{
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.02,
+  "bos_token_id": 50256,
+  "embd_pdrop": 0.02,
+  "eos_token_id": 50256,
+  "gradient_checkpointing": false,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_ctx": 1024,
+  "n_embd": 768,
+  "n_head": 12,
+  "n_inner": null,
+  "n_layer": 12,
+  "n_positions": 1024,
+  "pad_token_id": 50256,
+  "resid_pdrop": 0.02,
+  "scale_attn_weights": true,
+  "summary_activation": null,
+  "summary_first_dropout": 0.1,
+  "summary_proj_to_labels": true,
+  "summary_type": "cls_index",
+  "summary_use_proj": true,
+  "task_specific_params": {
+    "text-generation": {
+      "do_sample": true,
+      "max_length": 50
+    }
+  },
+  "transformers_version": "4.8.2",
+  "use_cache": true,
+  "vocab_size": 50257
+}

flax_model.msgpack

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:977e6a59716b48064a63c4acbbc4f3db504dc5051929c15d58253a242ddde376
+size 497764120

requirements.txt

@@ -0,0 +1,5 @@
+torch
+streamlit
+jax
+flax
+transformers