Update with h2oGPT hash 12617a0d8f901cf2e3de0601fbdb52b492b0bd2e

app.py

@@ -4,7 +4,7 @@ import sys
 import os
 import traceback
 import typing
-from utils import set_seed, flatten_list, clear_torch_cache, system_info_print, zip_data, save_generate_output
+from utils import set_seed, flatten_list, clear_torch_cache, system_info_print, zip_data, save_generate_output, s3up
 
 SEED = 1236
 set_seed(SEED)
@@ -891,25 +891,41 @@ body.dark{background:linear-gradient(#0d0d0d,#333333);}"""
                                     lora_used2 = gr.Textbox(label="Current LORA 2", value=no_lora_str,
                                                             visible=kwargs['show_lora'])
                 with gr.TabItem("System"):
+                    admin_row = gr.Row()
+                    with admin_row:
+                        admin_pass_textbox = gr.Textbox(label="Admin Password", type='password', visible=is_public)
+                        admin_btn = gr.Button(value="Admin Access", visible=is_public)
                     system_row = gr.Row(visible=not is_public)
-                    admin_pass_textbox = gr.Textbox(label="Admin Password", type='password', visible=is_public)
-                    admin_btn = gr.Button(value="admin", visible=is_public)
                     with system_row:
                         with gr.Column():
-                            system_text = gr.Textbox(label='System Info')
-                            system_btn = gr.Button(value='Get System Info')
-
-                            zip_btn = gr.Button("Zip")
-                            file_output = gr.File()
+                            with gr.Row():
+                                system_btn = gr.Button(value='Get System Info')
+                                system_text = gr.Textbox(label='System Info')
+
+                            with gr.Row():
+                                zip_btn = gr.Button("Zip")
+                                zip_text = gr.Textbox(label="Zip file name")
+                                file_output = gr.File()
+                            with gr.Row():
+                                s3up_btn = gr.Button("S3UP")
+                                s3up_text = gr.Textbox(label='S3UP result')
 
         # Get flagged data
         zip_data1 = functools.partial(zip_data, root_dirs=['flagged_data_points', kwargs['save_dir']])
-        zip_btn.click(zip_data1, inputs=None, outputs=file_output)
+        zip_btn.click(zip_data1, inputs=None, outputs=[file_output, zip_text])
+
+        #def update_s3(x):
+        #    return gr.update(value="S3UP [%s]" % x)
+        s3up_btn.click(s3up, inputs=zip_text, outputs=s3up_text)
 
         def check_admin_pass(x):
             return gr.update(visible=x == admin_pass)
 
-        admin_btn.click(check_admin_pass, inputs=admin_pass_textbox, outputs=system_row)
+        def close_admin(x):
+            return gr.update(visible=not (x == admin_pass))
+
+        admin_btn.click(check_admin_pass, inputs=admin_pass_textbox, outputs=system_row) \
+                 .then(close_admin, inputs=admin_pass_textbox, outputs=admin_row)
 
         # Get inputs to evaluate()
         inputs_list = get_inputs_list(locals(), kwargs['model_lower'])
@@ -1953,7 +1969,7 @@ if __name__ == "__main__":
     can also pass --prompt_type='human_bot' and model can somewhat handle instructions without being instruct tuned
     python generate.py --base_model=decapoda-research/llama-65b-hf --load_8bit=False --infer_devices=False --prompt_type='human_bot'
 
-    python generate.py --base_model=h2oai/h2ogpt-oig-oasst1-256-6.9b
+    python generate.py --base_model=h2oai/h2ogpt-oig-oasst1-512-6.9b
 
     """, flush=True)
     fire.Fire(main)

client_test.py

@@ -3,7 +3,7 @@ Client test.
 
 Run server:
 
-python generate.py  --base_model=h2oai/h2ogpt-oig-oasst1-256-6.9b
+python generate.py  --base_model=h2oai/h2ogpt-oig-oasst1-512-6.9b
 
 NOTE: For private models, add --use-auth_token=True
 

finetune.py

@@ -84,11 +84,9 @@ prompt_type_to_model_name = {
     'instruct_with_end': ['databricks/dolly-v2-12b'],
     'quality': [],
     'human_bot': [
-        'h2oai/h2ogpt-oig-oasst1-256-12b',
         'h2oai/h2ogpt-oasst1-512-12b',
-        'h2oai/h2ogpt-oasst1-256-20b',
         'h2oai/h2ogpt-oasst1-512-20b',
-        'h2oai/h2ogpt-oig-oasst1-256-6.9b',
+        'h2oai/h2ogpt-oig-oasst1-512-6.9b',
     ],
     'dai_faq': [],
     'summarize': [],

requirements.txt

@@ -26,6 +26,8 @@ tokenizers==0.13.3
 # optional for generate
 pynvml==11.5.0
 psutil==5.9.4
+boto3==1.26.101
+botocore==1.29.101
 
 # optional for finetune
 tensorboard==2.12.1

utils.py

@@ -118,7 +118,7 @@ def _zip_data(root_dirs=None, zip_file=None, base_dir='./'):
                     assert os.path.exists(file_to_archive)
                     path_to_archive = os.path.relpath(file_to_archive, base_dir)
                     expt_zip.write(filename=file_to_archive, arcname=path_to_archive)
-    return zip_file
+    return zip_file, zip_file
 
 
 def save_generate_output(output=None, base_model=None, save_dir=None):
@@ -152,3 +152,35 @@ def _save_generate_output(output=None, base_model=None, save_dir=None):
                     dict(text=output, time=time.ctime(), base_model=base_model)
                 ) + ",\n"
             )
+
+
+def s3up(filename):
+    try:
+        return _s3up(filename)
+    except Exception as e:
+        traceback.print_exc()
+        print('Exception for file %s in s3up: %s' % (filename, str(e)))
+        return "Failed to upload %s: Error: %s" % (filename, str(e))
+
+
+def _s3up(filename):
+    import boto3
+
+    aws_access_key_id = os.getenv('AWS_SERVER_PUBLIC_KEY')
+    aws_secret_access_key = os.getenv('AWS_SERVER_SECRET_KEY')
+    bucket = os.getenv('AWS_BUCKET')
+    assert aws_access_key_id, "Set AWS key"
+    assert aws_secret_access_key, "Set AWS secret"
+    assert bucket, "Set AWS Bucket"
+
+    s3 = boto3.client('s3',
+                      aws_access_key_id=os.getenv('AWS_SERVER_PUBLIC_KEY'),
+                      aws_secret_access_key=os.getenv('AWS_SERVER_SECRET_KEY'),
+                      )
+    ret = s3.upload_file(
+        Filename=filename,
+        Bucket=os.getenv('AWS_BUCKET'),
+        Key=filename,
+    )
+    if ret in [None, '']:
+        return "Successfully uploaded %s" % filename