Spaces:
Build error
Build error
jonathanagustin
commited on
Commit
•
76e3959
1
Parent(s):
a1d450d
Update app.py
Browse files
app.py
CHANGED
@@ -1,86 +1,14 @@
|
|
1 |
-
# Copyright 2024 X.AI Corp.
|
2 |
-
#
|
3 |
-
# Licensed under the Apache License, Version 2.0 (the "License");
|
4 |
-
# you may not use this file except in compliance with the License.
|
5 |
-
# You may obtain a copy of the License at
|
6 |
-
#
|
7 |
-
# http://www.apache.org/licenses/LICENSE-2.0
|
8 |
-
#
|
9 |
-
# Unless required by applicable law or agreed to in writing, software
|
10 |
-
# distributed under the License is distributed on an "AS IS" BASIS,
|
11 |
-
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
12 |
-
# See the License for the specific language governing permissions and
|
13 |
-
# limitations under the License.
|
14 |
-
|
15 |
-
import logging
|
16 |
-
import os
|
17 |
-
|
18 |
import gradio as gr
|
19 |
-
from
|
20 |
-
|
21 |
-
from model import LanguageModelConfig
|
22 |
-
from model import QuantizedWeight8bit as QW8Bit
|
23 |
-
from model import TransformerConfig
|
24 |
-
from runners import InferenceRunner, ModelRunner, sample_from_model
|
25 |
-
|
26 |
-
CKPT_PATH = "./checkpoints/"
|
27 |
-
|
28 |
-
|
29 |
-
def main():
|
30 |
-
grok_1_model = LanguageModelConfig(
|
31 |
-
vocab_size=128 * 1024,
|
32 |
-
pad_token=0,
|
33 |
-
eos_token=2,
|
34 |
-
sequence_len=8192,
|
35 |
-
embedding_init_scale=1.0,
|
36 |
-
output_multiplier_scale=0.5773502691896257,
|
37 |
-
embedding_multiplier_scale=78.38367176906169,
|
38 |
-
model=TransformerConfig(
|
39 |
-
emb_size=48 * 128,
|
40 |
-
widening_factor=8,
|
41 |
-
key_size=128,
|
42 |
-
num_q_heads=48,
|
43 |
-
num_kv_heads=8,
|
44 |
-
num_layers=64,
|
45 |
-
attn_output_multiplier=0.08838834764831845,
|
46 |
-
shard_activations=True,
|
47 |
-
# MoE.
|
48 |
-
num_experts=8,
|
49 |
-
num_selected_experts=2,
|
50 |
-
# Activation sharding.
|
51 |
-
data_axis="data",
|
52 |
-
model_axis="model",
|
53 |
-
),
|
54 |
-
)
|
55 |
-
inference_runner = InferenceRunner(
|
56 |
-
pad_sizes=(1024,),
|
57 |
-
runner=ModelRunner(
|
58 |
-
model=grok_1_model,
|
59 |
-
bs_per_device=0.125,
|
60 |
-
checkpoint_path=CKPT_PATH,
|
61 |
-
),
|
62 |
-
name="local",
|
63 |
-
load=CKPT_PATH,
|
64 |
-
tokenizer_path="./tokenizer.model",
|
65 |
-
local_mesh_config=(1, 8),
|
66 |
-
between_hosts_config=(1, 1),
|
67 |
-
)
|
68 |
-
|
69 |
-
inference_runner.initialize()
|
70 |
-
gen = inference_runner.run()
|
71 |
-
|
72 |
-
def chatbot_response(message, history):
|
73 |
-
history.append(message)
|
74 |
-
inp = " ".join(history)
|
75 |
-
output = sample_from_model(inference_runner.run(), inp, max_len=100, temperature=0.01)
|
76 |
-
history.append(output)
|
77 |
-
return output
|
78 |
-
|
79 |
-
iface = gr.ChatInterface(fn=chatbot_response, chatbot="Grok-1", history=[], title="Grok-1")
|
80 |
|
81 |
-
|
82 |
|
|
|
|
|
|
|
|
|
|
|
|
|
83 |
|
84 |
-
|
85 |
-
|
86 |
-
main()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import gradio as gr
|
2 |
+
from transformers import pipeline
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
|
4 |
+
pipe = pipeline("text-generation", model="keyfan/grok-1-hf", trust_remote_code=True)
|
5 |
|
6 |
+
def chatbot_response(message, history):
|
7 |
+
history.append(message)
|
8 |
+
inp = " ".join(history)
|
9 |
+
output = pipe(inp, max_length=100, temperature=0.01)[0]['generated_text'] # Access generated text
|
10 |
+
history.append(output)
|
11 |
+
return output
|
12 |
|
13 |
+
iface = gr.ChatInterface(fn=chatbot_response, chatbot="Grok-1", history=[], title="Grok-1")
|
14 |
+
iface.launch()
|
|