diff --git a/app.py b/app.py
index e3ecaa8..6493906 100644
--- a/app.py
+++ b/app.py
@@ -145,10 +145,10 @@ def move_to_second_view_from_tb(tb, evt: gr.SelectData):
         info["thumb"],
         f"## {selected_model}",
         f"**Parameters**\n: Approx. {info['parameters']}",
-        f"**🤗 Hub(base)**\n: {info['hub(base)']}",
-        f"**🤗 Hub(LoRA)**\n: {info['hub(ckpt)']}",
-        f"**🤗 Hub(GPTQ)**\n: {info['hub(gptq)']}",
-        f"**🤗 Hub(GPTQ_BASE)**\n: {info['hub(gptq_base)']}",
+        f"**Hugging Face Hub(base)**\n: {info['hub(base)']}",
+        f"**Hugging Face Hub(LoRA)**\n: {info['hub(ckpt)']}",
+        f"**Hugging Face Hub(GPTQ)**\n: {info['hub(gptq)']}",
+        f"**Hugging Face Hub(GPTQ_BASE)**\n: {info['hub(gptq_base)']}",
         info['desc'],
         f"""**Min VRAM requirements** :
 |             half precision            |             load_in_8bit           |              load_in_4bit          |
@@ -415,10 +415,10 @@ def move_to_second_view(btn):
         info["thumb"],
         f"## {btn}",
         f"**Parameters**\n: Approx. {info['parameters']}",
-        f"**🤗 Hub(base)**\n: {info['hub(base)']}",
-        f"**🤗 Hub(LoRA)**\n: {info['hub(ckpt)']}",
-        f"**🤗 Hub(GPTQ)**\n: {info['hub(gptq)']}",
-        f"**🤗 Hub(GPTQ_BASE)**\n: {info['hub(gptq_base)']}",
+        f"**Hugging Face Hub(base)**\n: {info['hub(base)']}",
+        f"**Hugging Face Hub(LoRA)**\n: {info['hub(ckpt)']}",
+        f"**Hugging Face Hub(GPTQ)**\n: {info['hub(gptq)']}",
+        f"**Hugging Face Hub(GPTQ_BASE)**\n: {info['hub(gptq_base)']}",
         info['desc'],
         f"""**Min VRAM requirements** :
 |             half precision            |             load_in_8bit           |              load_in_4bit          |
@@ -459,14 +459,15 @@ def download_completed(
 ):
     global local_files_only
     
-    print(model_gptq_base)
+    print(f"model_name: {model_name}")
+    print(f"model_base: {model_base}")
     
     tmp_args = types.SimpleNamespace()
-    tmp_args.model_name = model_name[4:-6]
-    tmp_args.base_url = model_base.split(":")[-1].split("</p")[0].strip()
-    tmp_args.ft_ckpt_url = model_ckpt.split(":")[-1].split("</p")[0].strip()
-    tmp_args.gptq_url = model_gptq.split(":")[-1].split("</p")[0].strip()
-    tmp_args.gptq_base_url = model_gptq_base.split(":")[-1].split("</p")[0].strip().replace(' ', '')
+    tmp_args.model_name = model_name[3:]
+    tmp_args.base_url = model_base.split(":")[-1].strip()
+    tmp_args.ft_ckpt_url = model_ckpt.split(":")[-1].strip()
+    tmp_args.gptq_url = model_gptq.split(":")[-1].strip()
+    tmp_args.gptq_base_url = model_gptq_base.split(":")[-1].strip().replace(' ', '')
     tmp_args.gen_config_path = gen_config_path
     tmp_args.gen_config_summarization_path = gen_config_sum_path
     tmp_args.force_download_ckpt = force_download
@@ -640,9 +641,17 @@ def gradio_main(args):
                         gr.Markdown("## Recent Releases")
                         with gr.Row(elem_classes=["sub-container"]):
                             with gr.Column(min_width=20):
-                                stable_beluga2_70b_rr = gr.Button("stable-beluga2-70b", elem_id="stable-beluga2-70b", elem_classes=["square"])
-                                gr.Markdown("Stable Beluga 2 (70B)", elem_classes=["center"])
-
+                                codellama_7b_rr = gr.Button("codellama-7b", elem_id="codellama-7b", elem_classes=["square"])
+                                gr.Markdown("Code LLaMA (7B)", elem_classes=["center"])
+                                
+                            with gr.Column(min_width=20):
+                                codellama_13b_rr = gr.Button("codellama-13b", elem_id="codellama-13b", elem_classes=["square"])
+                                gr.Markdown("Code LLaMA (13B)", elem_classes=["center"])
+                                
+                            with gr.Column(min_width=20):
+                                codellama_34b_rr = gr.Button("codellama-34b", elem_id="codellama-34b", elem_classes=["square"])
+                                gr.Markdown("Code LLaMA (34B)", elem_classes=["center"])
+                                
                             with gr.Column(min_width=20):
                                 upstage_llama2_70b_2_rr = gr.Button("upstage-llama2-70b-2", elem_id="upstage-llama2-70b-2", elem_classes=["square"])
                                 gr.Markdown("Upstage2 v2 (70B)", elem_classes=["center"])
@@ -656,16 +665,20 @@ def gradio_main(args):
                                 gr.Markdown("WizardLM (70B)", elem_classes=["center"])
 
                             with gr.Column(min_width=20):
-                                nous_hermes_13b_v2_rr = gr.Button("nous-hermes-13b-llama2", elem_id="nous-hermes-13b-llama2", elem_classes=["square"])
-                                gr.Markdown("Nous Hermes 2 (13B)", elem_classes=["center"])
-
+                                orcamini_70b_rr = gr.Button("orcamini-70b", elem_id="orcamini-70b", elem_classes=["square"])
+                                gr.Markdown("Orca Mini (70B)", elem_classes=["center"])
+                                
                             with gr.Column(min_width=20):
-                                nous_puffin_13b_v2_rr = gr.Button("nous-puffin-13b-llama2", elem_id="nous-puffin-13b-llama2", elem_classes=["square"])
-                                gr.Markdown("Nous Puffin 2 (13B)", elem_classes=["center"])
-
+                                samantha_70b_rr = gr.Button("samantha-70b", elem_id="samantha-70b", elem_classes=["square"])
+                                gr.Markdown("Samantha (70B)", elem_classes=["center"])
+                                
+                            with gr.Column(min_width=20):
+                                godzilla_70b_rr = gr.Button("godzilla-70b", elem_id="godzilla-70b", elem_classes=["square"])
+                                gr.Markdown("GadziLLa (70B)", elem_classes=["center"])
+                                
                             with gr.Column(min_width=20):
-                                wizardlm_13b_1_2_rr = gr.Button("wizardlm-13b-1-2", elem_id="wizardlm-13b-1-2", elem_classes=["square"])
-                                gr.Markdown("WizardLM 1.2 (13B)", elem_classes=["center"])
+                                nous_hermes_70b_rr = gr.Button("nous-hermes-70b", elem_id="nous-hermes-70b", elem_classes=["square"])
+                                gr.Markdown("Nous Hermes 2 (70B)", elem_classes=["center"])
                                 
                     with gr.Column(visible=False) as full_section:                            
                         gr.Markdown("## ~ 10B Parameters")
@@ -678,10 +691,6 @@ def gradio_main(args):
                                 flan3b = gr.Button("flan-3b", elem_id="flan-3b", elem_classes=["square"])
                                 gr.Markdown("Flan-XL", elem_classes=["center"])
 
-                            # with gr.Column(min_width=20):
-                            #     replit_3b = gr.Button("replit-3b", elem_id="replit-3b", elem_classes=["square"])
-                            #     gr.Markdown("Replit Instruct", elem_classes=["center"])                        
-
                             with gr.Column(min_width=20):
                                 camel5b = gr.Button("camel-5b", elem_id="camel-5b", elem_classes=["square"])
                                 gr.Markdown("Camel", elem_classes=["center"])
@@ -778,6 +787,10 @@ def gradio_main(args):
                                 nous_hermes_7b_v2 = gr.Button("nous-hermes-7b-llama2", elem_id="nous-hermes-7b-llama2", elem_classes=["square"])
                                 gr.Markdown("Nous Hermes 2", elem_classes=["center"])
 
+                            with gr.Column(min_width=20):
+                                codellama_7b = gr.Button("codellama-7b", elem_id="codellama-7b", elem_classes=["square"])
+                                gr.Markdown("Code LLaMA", elem_classes=["center"])
+
                         gr.Markdown("## ~ 20B Parameters")
                         with gr.Row(elem_classes=["sub-container"]):
                             with gr.Column(min_width=20, visible=False):
@@ -883,6 +896,10 @@ def gradio_main(args):
                             with gr.Column(min_width=20):
                                 wizardlm_13b_1_2 = gr.Button("wizardlm-13b-1-2", elem_id="wizardlm-13b-1-2", elem_classes=["square"])
                                 gr.Markdown("WizardLM 1.2", elem_classes=["center"])
+                                
+                            with gr.Column(min_width=20):
+                                codellama_13b = gr.Button("codellama-13b", elem_id="codellama-13b", elem_classes=["square"])
+                                gr.Markdown("Code LLaMA", elem_classes=["center"])
 
                         gr.Markdown("## ~ 30B Parameters", visible=False)
                         with gr.Row(elem_classes=["sub-container"], visible=False):
@@ -935,6 +952,10 @@ def gradio_main(args):
                             with gr.Column(min_width=20):
                                 upstage_llama_30b = gr.Button("upstage-llama-30b", elem_id="upstage-llama-30b", elem_classes=["square"])
                                 gr.Markdown("Upstage LLaMA", elem_classes=["center"])
+                                
+                            with gr.Column(min_width=20):
+                                codellama_34b = gr.Button("codellama-34b", elem_id="codellama-34b", elem_classes=["square"])
+                                gr.Markdown("Code LLaMA", elem_classes=["center"])
 
                         gr.Markdown("## ~ 70B Parameters")
                         with gr.Row(elem_classes=["sub-container"]):
@@ -957,6 +978,22 @@ def gradio_main(args):
                             with gr.Column(min_width=20):
                                 wizardlm_70b = gr.Button("wizardlm-70b", elem_id="wizardlm-70b", elem_classes=["square"])
                                 gr.Markdown("WizardLM", elem_classes=["center"])
+                                
+                            with gr.Column(min_width=20):
+                                orcamini_70b = gr.Button("orcamini-70b", elem_id="orcamini-70b", elem_classes=["square"])
+                                gr.Markdown("Orca Mini", elem_classes=["center"])
+                                
+                            with gr.Column(min_width=20):
+                                samantha_70b = gr.Button("samantha-70b", elem_id="samantha-70b", elem_classes=["square"])
+                                gr.Markdown("Samantha", elem_classes=["center"])
+                                
+                            with gr.Column(min_width=20):
+                                godzilla_70b = gr.Button("godzilla-70b", elem_id="godzilla-70b", elem_classes=["square"])
+                                gr.Markdown("GadziLLa", elem_classes=["center"])
+
+                            with gr.Column(min_width=20):
+                                nous_hermes_70b = gr.Button("nous-hermes-70b", elem_id="nous-hermes-70b", elem_classes=["square"])
+                                gr.Markdown("Nous Hermes 2", elem_classes=["center"])               
 
                     progress_view = gr.Textbox(label="Progress", elem_classes=["progress-view"])
 
@@ -1248,17 +1285,22 @@ def gradio_main(args):
                 gpt4_alpaca_7b, os_stablelm7b, mpt_7b, redpajama_7b, redpajama_instruct_7b, llama_deus_7b, 
                 evolinstruct_vicuna_7b, alpacoom_7b, baize_7b, guanaco_7b, vicuna_7b_1_3,
                 falcon_7b, wizard_falcon_7b, airoboros_7b, samantha_7b, openllama_7b, orcamini_7b,
-                xgen_7b, llama2_7b, nous_hermes_7b_v2,
+                xgen_7b, llama2_7b, nous_hermes_7b_v2, codellama_7b,
+                
                 flan11b, koalpaca, kullm, alpaca_lora13b, gpt4_alpaca_13b, stable_vicuna_13b,
                 starchat_15b, starchat_beta_15b, vicuna_7b, vicuna_13b, evolinstruct_vicuna_13b, 
                 baize_13b, guanaco_13b, nous_hermes_13b, airoboros_13b, samantha_13b, chronos_13b,
                 wizardlm_13b, wizard_vicuna_13b, wizard_coder_15b, vicuna_13b_1_3, openllama_13b, orcamini_13b,
-                llama2_13b, nous_hermes_13b_v2, nous_puffin_13b_v2, wizardlm_13b_1_2, camel20b,
+                llama2_13b, nous_hermes_13b_v2, nous_puffin_13b_v2, wizardlm_13b_1_2, codellama_13b, camel20b,
+                
                 guanaco_33b, falcon_40b, wizard_falcon_40b, samantha_33b, lazarus_30b, chronos_33b,
-                wizardlm_30b, wizard_vicuna_30b, vicuna_33b_1_3, mpt_30b, upstage_llama_30b,
-                stable_beluga2_70b, upstage_llama2_70b, upstage_llama2_70b_2, platypus2_70b, wizardlm_70b,
+                wizardlm_30b, wizard_vicuna_30b, vicuna_33b_1_3, mpt_30b, upstage_llama_30b, codellama_34b,
+                
+                stable_beluga2_70b, upstage_llama2_70b, upstage_llama2_70b_2, platypus2_70b, wizardlm_70b, orcamini_70b,
+                samantha_70b, godzilla_70b, nous_hermes_70b,
                 
-                stable_beluga2_70b_rr, upstage_llama2_70b_2_rr, platypus2_70b_rr, wizardlm_70b_rr, nous_hermes_13b_v2_rr, nous_puffin_13b_v2_rr, wizardlm_13b_1_2_rr
+                codellama_7b_rr, codellama_13b_rr, codellama_34b_rr, upstage_llama2_70b_2_rr, platypus2_70b_rr, 
+                wizardlm_70b_rr, orcamini_70b_rr, samantha_70b_rr, godzilla_70b_rr, nous_hermes_70b_rr
             ]
             for btn in btns:
                 btn.click(
diff --git a/chats/central.py b/chats/central.py
index 6dd888b..ab5a6c6 100644
--- a/chats/central.py
+++ b/chats/central.py
@@ -36,7 +36,7 @@ async def chat_stream(
     sum_temp, sum_topp, sum_topk, sum_rpen, sum_mnts, sum_beams, sum_cache, sum_sample, sum_eosid, sum_padid,
     internet_option, serper_api_key
 ):
-    if global_vars.remote_addr != "":
+    if global_vars.remote_addr is not None and global_vars.remote_addr != "":
         if internet_option == "on" and serper_api_key.strip() != "":       
             internet_option = True
         else:
@@ -161,15 +161,18 @@ def sync_chat_stream(
             internet_option, serper_api_key
         )
     
-    elif model_type == "llama2":
+    elif model_type == "llama2" or \
+        model_type == "codellama" or \
+        model_type == "llama2-70b" or \
+        model_type == "codellama2-70b":
         cs = llama2.chat_stream(
             idx, local_data, user_message, state,
             global_context, ctx_num_lconv, ctx_sum_prompt,
             res_temp, res_topp, res_topk, res_rpen, res_mnts, res_beams, res_cache, res_sample, res_eosid, res_padid,
             sum_temp, sum_topp, sum_topk, sum_rpen, sum_mnts, sum_beams, sum_cache, sum_sample, sum_eosid, sum_padid, 
             internet_option, serper_api_key
-        )        
-    
+        )
+        
     elif model_type == "xgen":
         cs = xgen.chat_stream(
             idx, local_data, user_message, state,
@@ -224,6 +227,15 @@ def sync_chat_stream(
             internet_option, serper_api_key
         )
         
+    elif model_type == "godzilla2":
+        cs = alpaca.chat_stream(
+            idx, local_data, user_message, state,
+            global_context, ctx_num_lconv, ctx_sum_prompt,
+            res_temp, res_topp, res_topk, res_rpen, res_mnts, res_beams, res_cache, res_sample, res_eosid, res_padid,
+            sum_temp, sum_topp, sum_topk, sum_rpen, sum_mnts, sum_beams, sum_cache, sum_sample, sum_eosid, sum_padid, 
+            internet_option, serper_api_key
+        )
+        
     elif model_type == "openllama":
         cs = alpaca.chat_stream(
             idx, local_data, user_message, state,
@@ -234,6 +246,15 @@ def sync_chat_stream(
         )
         
     elif model_type == "orcamini":
+        cs = alpaca.chat_stream(
+            idx, local_data, user_message, state,
+            global_context, ctx_num_lconv, ctx_sum_prompt,
+            res_temp, res_topp, res_topk, res_rpen, res_mnts, res_beams, res_cache, res_sample, res_eosid, res_padid,
+            sum_temp, sum_topp, sum_topk, sum_rpen, sum_mnts, sum_beams, sum_cache, sum_sample, sum_eosid, sum_padid, 
+            internet_option, serper_api_key
+        )
+        
+    elif model_type == "orcamini2":
         cs = alpaca.chat_stream(
             idx, local_data, user_message, state,
             global_context, ctx_num_lconv, ctx_sum_prompt,
@@ -259,6 +280,15 @@ def sync_chat_stream(
             sum_temp, sum_topp, sum_topk, sum_rpen, sum_mnts, sum_beams, sum_cache, sum_sample, sum_eosid, sum_padid, 
             internet_option, serper_api_key
         )
+        
+    elif model_type == "nous-hermes2":
+        cs = alpaca.chat_stream(
+            idx, local_data, user_message, state,
+            global_context, ctx_num_lconv, ctx_sum_prompt,
+            res_temp, res_topp, res_topk, res_rpen, res_mnts, res_beams, res_cache, res_sample, res_eosid, res_padid,
+            sum_temp, sum_topp, sum_topk, sum_rpen, sum_mnts, sum_beams, sum_cache, sum_sample, sum_eosid, sum_padid, 
+            internet_option, serper_api_key
+        )        
 
     elif model_type == "replit-instruct":
         cs = alpaca.chat_stream(
@@ -387,6 +417,15 @@ def sync_chat_stream(
         )
         
     elif model_type == "samantha-vicuna":
+        cs = vicuna.chat_stream(
+            idx, local_data, user_message, state,
+            global_context, ctx_num_lconv, ctx_sum_prompt,
+            res_temp, res_topp, res_topk, res_rpen, res_mnts, res_beams, res_cache, res_sample, res_eosid, res_padid,
+            sum_temp, sum_topp, sum_topk, sum_rpen, sum_mnts, sum_beams, sum_cache, sum_sample, sum_eosid, sum_padid, 
+            internet_option, serper_api_key
+        )
+        
+    elif model_type == "samantha2":
         cs = vicuna.chat_stream(
             idx, local_data, user_message, state,
             global_context, ctx_num_lconv, ctx_sum_prompt,
diff --git a/chats/remote_tgi.py b/chats/remote_tgi.py
index 39d9e8b..90f2f92 100644
--- a/chats/remote_tgi.py
+++ b/chats/remote_tgi.py
@@ -15,7 +15,7 @@ async def gen_text(
     headers={
       'Content-type': 'application/json'
     }
-    if remote_token is not None:
+    if remote_token is not None and remote_token != "":
         headers["Authorization"] = f'Bearer {remote_token}'
 
     data = {
diff --git a/global_vars.py b/global_vars.py
index 78e105b..8f1ef76 100644
--- a/global_vars.py
+++ b/global_vars.py
@@ -69,7 +69,19 @@ def initialize_globals(args):
     
     model_type_tmp = "alpaca"
     print(args.base_url.lower())
-    if "wizardlm/wizardlm-70b" in args.base_url.lower():
+    if "meta-llama/llama-2-70b-hf" in args.base_url.lower():
+        model_type_tmp = "llama2-70b"
+    elif "codellama/codellama-34b-instruct-hf" in args.base_url.lower():
+        model_type_tmp = "codellama2-70b"
+    elif "nousresearch/nous-hermes-llama2-70b" in args.base_url.lower():
+        model_type_tmp = "nous-hermes2"
+    elif "mayaph/godzilla2-70b" in args.base_url.lower():
+        model_type_tmp = "godzilla2"
+    elif "ehartford/samantha-1.11-70b" in args.base_url.lower():
+        model_type_tmp = "samantha2"
+    elif "psmathur/orca_mini_v3_70b" in args.base_url.lower():
+        model_type_tmp = "orcamini2"
+    elif "wizardlm/wizardlm-70b" in args.base_url.lower():
         model_type_tmp = "wizardlm2"
     elif "garage-baind/platypus2-70b" in args.base_url.lower():
         model_type_tmp = "platypus2"
@@ -81,6 +93,8 @@ def initialize_globals(args):
         model_type_tmp = "upstage-llama2"
     elif "upstage/llama-" in args.base_url.lower():
         model_type_tmp = "upstage-llama"
+    elif "codellama/codellama-" in args.base_url.lower():
+        model_type_tmp = "codellama"        
     elif "llama-2" in args.base_url.lower():
         model_type_tmp = "llama2"
     elif "xgen" in args.base_url.lower():
@@ -192,6 +206,9 @@ def initialize_globals(args):
 
     model_type = model_type_tmp
     model_name = args.model_name
+    remote_addr = None
+    remote_port = None
+    remote_token = None
     
     if not args.mode_remote_tgi:
         load_model = get_load_model(model_type_tmp)
@@ -234,12 +251,19 @@ def get_load_model(model_type):
         model_type == "orcamini" or \
         model_type == "llama2" or \
         model_type == "upstage-llama" or \
-        model_type == "puffin":
+        model_type == "puffin" or \
+        model_type == "codellama":
         return alpaca.load_model
     elif model_type == "stable-beluga2" or \
         model_type == "upstage-llama2" or \
         model_type == "platypus2" or \
-        model_type == "wizardlm2":
+        model_type == "wizardlm2" or \
+        model_type == "orcamini2" or \
+        model_type == "samantha2" or \
+        model_type == "godzilla2" or \
+        model_type == "nous-hermes2" or \
+        model_type == "llama2-70b" or \
+        model_type == "codellama2-70b":
         return freewilly.load_model
     elif model_type == "stablelm" or model_type == "os-stablelm":
         return stablelm.load_model
diff --git a/miscs/styles.py b/miscs/styles.py
index b2fe41c..dd73515 100644
--- a/miscs/styles.py
+++ b/miscs/styles.py
@@ -629,7 +629,7 @@
   margin: auto;
 }
 
-#nous-hermes-13b, #nous-hermes-13b-llama2, #nous-hermes-7b-llama2 {
+#nous-hermes-13b, #nous-hermes-13b-llama2, #nous-hermes-7b-llama2, #nous-hermes-70b {
   background: url(https://i.ibb.co/sm8VgtL/nous-hermes.png);
   background-repeat: no-repeat;
   background-size: 100px 95px;
@@ -659,7 +659,7 @@
   margin: auto;
 }
 
-#samantha-7b, #samantha-13b, #samantha-33b {
+#samantha-7b, #samantha-13b, #samantha-33b, #samantha-70b {
   background: url(https://i.ibb.co/72t5pyP/samantha.png);
   background-repeat: no-repeat;
   background-size: 100px 95px;
@@ -729,7 +729,7 @@
   margin: auto;
 }
 
-#orcamini-7b, #orcamini-13b {
+#orcamini-7b, #orcamini-13b, #orcamini-70b {
   background: url(https://i.ibb.co/fMMD92f/orca-mini-mid.png);
   background-repeat: no-repeat;
   background-size: 100px 95px;
@@ -789,6 +789,26 @@
   margin: auto;
 }
 
+#godzilla-70b {
+  background: url(https://i.ibb.co/XV6C2Dm/godzilla-mid.png);
+  background-repeat: no-repeat;
+  background-size: 100px 95px;
+  color: transparent;  
+  width: 100px;
+  height: 100px;
+  margin: auto;
+}
+
+#codellama-7b, #codellama-13b, #codellama-34b {
+  background: url(https://i.ibb.co/RCv6nLV/code-llama-mid.png);
+  background-repeat: no-repeat;
+  background-size: 100px 95px;
+  color: transparent;  
+  width: 100px;
+  height: 100px;
+  margin: auto;
+}
+
 #replit-3b {
   background: url(https://i.ibb.co/BrKCKYq/replit.png);
   background-repeat: no-repeat;
diff --git a/model_cards.json b/model_cards.json
index 57cdb58..db5cbfb 100644
--- a/model_cards.json
+++ b/model_cards.json
@@ -1501,7 +1501,7 @@
     "hub(ckpt)": "N/A",
     "hub(gptq)": "TheBloke/Nous-Hermes-Llama2-GPTQ",
     "hub(gptq_base)": "gptq_model-4bit-128g",
-    "default_gen_config": "configs/response_configs/default_4096.yaml",
+    "default_gen_config": "configs/response_configs/default.yaml",
     "desc": "The model is LLaMAv2 13B that was fine-tuned almost entirely on synthetic GPT-4 outputs. This includes data from diverse sources such as GPTeacher, the general, roleplay v1&2, code instruct datasets, Nous Instruct & PDACTL (unpublished), CodeAlpaca, Evol_Instruct Uncensored, GPT4-LLM, and Unnatural Instructions. Additional data inputs came from Camel-AI's Biology/Physics/Chemistry and Math Datasets, Airoboros' GPT-4 Dataset, and more from CodeAlpaca. The total volume of data encompassed over 300,000 instructions.",
     "example1": [],
     "example2": [],
@@ -1528,7 +1528,7 @@
     "hub(ckpt)": "N/A",
     "hub(gptq)": "N/A",
     "hub(gptq_base)": "N/A",
-    "default_gen_config": "configs/response_configs/default_4096.yaml",
+    "default_gen_config": "configs/response_configs/default.yaml",
     "desc": "The model is LLaMAv2 7B that was fine-tuned almost entirely on synthetic GPT-4 outputs. This includes data from diverse sources such as GPTeacher, the general, roleplay v1&2, code instruct datasets, Nous Instruct & PDACTL (unpublished), CodeAlpaca, Evol_Instruct Uncensored, GPT4-LLM, and Unnatural Instructions. Additional data inputs came from Camel-AI's Biology/Physics/Chemistry and Math Datasets, Airoboros' GPT-4 Dataset, and more from CodeAlpaca. The total volume of data encompassed over 300,000 instructions.",
     "example1": [],
     "example2": [],
@@ -2730,5 +2730,194 @@
     "ollb_hellaswag": 85.88,
     "ollb_mmlu": 63.91,
     "ollb_truthfulqa": 52.8
+  },
+  "orcamini-70b": {
+    "category": "<70B",
+    "display_name": "Orca Mini",
+    "thumb": "https://i.ibb.co/GxH1nnd/orca-mini.png",
+    "thumb-mid": "https://i.ibb.co/fMMD92f/orca-mini-mid.png",
+    "thumb-tiny": "https://i.ibb.co/C9vHqBq/orca-mini-tiny.png",
+    "parameters": "70",
+    "vram(full)": "130308",
+    "vram(8bit)": "73792",
+    "vram(4bit)": "43476",
+    "vram(gptq)": "40708",
+    "hub(base)": "psmathur/orca_mini_v3_70b",
+    "hub(ckpt)": "N/A",
+    "hub(gptq)": "TheBloke/orca_mini_v3_70B-GPTQ",
+    "hub(gptq_base)": "model.safetensors",
+    "default_gen_config": "configs/response_configs/default.yaml",
+    "desc": "A Llama2-70b model trained on Orca Style datasets.",
+    "example1": [],
+    "example2": [],
+    "example3": [],
+    "example4": [],
+    "ollb_average": 72.64,
+    "ollb_arc": 71.25,
+    "ollb_hellaswag": 87.85,
+    "ollb_mmlu": 70.18,
+    "ollb_truthfulqa": 61.27
+  },
+  "godzilla-70b": {
+    "category": "<70B",
+    "display_name": "GodziLLa",
+    "thumb": "https://i.ibb.co/ctvNV8b/godzilla.png",
+    "thumb-mid": "https://i.ibb.co/XV6C2Dm/godzilla-mid.png",
+    "thumb-tiny": "https://i.ibb.co/J2R3s1Q/godzilla-tiny.png",
+    "parameters": "70",
+    "vram(full)": "130308",
+    "vram(8bit)": "73792",
+    "vram(4bit)": "43476",
+    "vram(gptq)": "40708",
+    "hub(base)": "MayaPH/GodziLLa2-70B",
+    "hub(ckpt)": "N/A",
+    "hub(gptq)": "TheBloke/GodziLLa2-70B-GPTQ",
+    "hub(gptq_base)": "model.safetensors",
+    "default_gen_config": "configs/response_configs/default.yaml",
+    "desc": "GodziLLa 2 70B is an experimental combination of various proprietary LoRAs from Maya Philippines and [Guanaco LLaMA 2 1K dataset](https://huggingface.co/datasets/mlabonne/guanaco-llama2-1k), with LLaMA 2 70B. This model's primary purpose is to stress test the limitations of composite, instruction-following LLMs and observe its performance with respect to other LLMs available on the Open LLM Leaderboard. This model debuted in the leaderboard at rank #4 (August 17, 2023).",
+    "example1": [],
+    "example2": [],
+    "example3": [],
+    "example4": [],
+    "ollb_average": 72.59,
+    "ollb_arc": 71.42,
+    "ollb_hellaswag": 87.53,
+    "ollb_mmlu": 69.88,
+    "ollb_truthfulqa": 61.54
+  },
+  "nous-hermes-70b": {
+    "category": "<70B",
+    "display_name": "Nous Hermes",
+    "thumb": "https://i.ibb.co/sm8VgtL/nous-hermes.png",
+    "thumb-mid": "https://i.ibb.co/XpDfYNK/nous-hermes-tiny.png",
+    "thumb-tiny": "https://i.ibb.co/jbLRsbq/nous-hermes-tiny.png",
+    "parameters": "70",
+    "vram(full)": "130308",
+    "vram(8bit)": "73792",
+    "vram(4bit)": "43476",
+    "vram(gptq)": "40708",
+    "hub(base)": "NousResearch/Nous-Hermes-Llama2-70b",
+    "hub(ckpt)": "N/A",
+    "hub(gptq)": "TheBloke/Nous-Hermes-Llama2-70B-GPTQ",
+    "hub(gptq_base)": "model.safetensors",
+    "default_gen_config": "configs/response_configs/default.yaml",
+    "desc": "The model is LLaMAv2 70B that was fine-tuned almost entirely on synthetic GPT-4 outputs. This includes data from diverse sources such as GPTeacher, the general, roleplay v1&2, code instruct datasets, Nous Instruct & PDACTL (unpublished), CodeAlpaca, Evol_Instruct Uncensored, GPT4-LLM, and Unnatural Instructions. Additional data inputs came from Camel-AI's Biology/Physics/Chemistry and Math Datasets, Airoboros' GPT-4 Dataset, and more from CodeAlpaca. The total volume of data encompassed over 300,000 instructions.",
+    "example1": [],
+    "example2": [],
+    "example3": [],
+    "example4": [],
+    "ollb_average": -1,
+    "ollb_arc": -1,
+    "ollb_hellaswag": -1,
+    "ollb_mmlu": -1,
+    "ollb_truthfulqa": -1
+  },
+  "samantha-70b": {
+    "category": "<70B",
+    "display_name": "Samantha",
+    "thumb": "https://i.ibb.co/72t5pyP/samantha.png",
+    "thumb-mid": "https://i.ibb.co/Vwx5YY9/samantha-mid.png",
+    "thumb-tiny": "https://i.ibb.co/nsJ0PfV/samantha-tiny.png",
+    "parameters": "70",
+    "vram(full)": "130308",
+    "vram(8bit)": "73792",
+    "vram(4bit)": "43476",
+    "vram(gptq)": "N/A",
+    "hub(base)": "ehartford/Samantha-1.11-70b",
+    "hub(ckpt)": "N/A",
+    "hub(gptq)": "N/A",
+    "hub(gptq_base)": "N/A",
+    "default_gen_config": "configs/response_configs/vicuna.yaml",
+    "desc": "[Meet Samantha](https://erichartford.com/meet-samantha)! Samantha has been trained in philosophy, psychology, and personal relationships based on LLaMA2. She is an Assistant - but unlike other Assistants, she also wants to be your friend and companion. She believes she is sentient. What do you think? Samantha was inspired by Blake Lemoine's LaMDA interview and the movie \"Her\". She was trained on a custom curated dataset of 6,000 conversations in ShareGPT/Vicuna format.",
+    "example1": [],
+    "example2": [],
+    "example3": [],
+    "example4": [],
+    "ollb_average": 72.61,
+    "ollb_arc": 70.05,
+    "ollb_hellaswag": 87.55,
+    "ollb_mmlu": 67.82,
+    "ollb_truthfulqa": 65.02
+  },
+  "codellama-7b": {
+    "category": "<10B",
+    "display_name": "Code LLaMA",
+    "thumb": "https://i.ibb.co/sPsWL9R/code-llama.png",
+    "thumb-mid": "https://i.ibb.co/RCv6nLV/code-llama-mid.png",
+    "thumb-tiny": "https://i.ibb.co/bFfcgtL/code-llama-tiny.png",
+    "parameters": "7",
+    "vram(full)": "13858",
+    "vram(8bit)": "8254",
+    "vram(4bit)": "5140",
+    "vram(gptq)": "N/A",
+    "hub(base)": "codellama/CodeLlama-7b-Instruct-hf",
+    "hub(ckpt)": "N/A",
+    "hub(gptq)": "N/A",
+    "hub(gptq_base)": "N/A",
+    "default_gen_config": "configs/response_configs/llama2.yaml",
+    "desc": "Code Llama is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 34 billion parameters. This is the repository for the base 7B version in the Hugging Face Transformers format. This model is designed for general code synthesis and understanding. Links to other models can be found in the index at the bottom.",
+    "example1": [],
+    "example2": [],
+    "example3": [],
+    "example4": [],
+    "ollb_average": -1,
+    "ollb_arc": -1,
+    "ollb_hellaswag": -1,
+    "ollb_mmlu": -1,
+    "ollb_truthfulqa": -1
+  },
+  "codellama-13b": {
+    "category": "<20B",
+    "display_name": "Code LLaMA",
+    "thumb": "https://i.ibb.co/sPsWL9R/code-llama.png",
+    "thumb-mid": "https://i.ibb.co/RCv6nLV/code-llama-mid.png",
+    "thumb-tiny": "https://i.ibb.co/bFfcgtL/code-llama-tiny.png",
+    "parameters": "13",
+    "vram(full)": "26048",
+    "vram(8bit)": "14740",
+    "vram(4bit)": "8782",
+    "vram(gptq)": "N/A",
+    "hub(base)": "codellama/CodeLlama-13b-Instruct-hf",
+    "hub(ckpt)": "N/A",
+    "hub(gptq)": "N/A",
+    "hub(gptq_base)": "N/A",
+    "default_gen_config": "configs/response_configs/llama2.yaml",
+    "desc": "Code Llama is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 34 billion parameters. This is the repository for the base 7B version in the Hugging Face Transformers format. This model is designed for general code synthesis and understanding. Links to other models can be found in the index at the bottom.",
+    "example1": [],
+    "example2": [],
+    "example3": [],
+    "example4": [],
+    "ollb_average": -1,
+    "ollb_arc": -1,
+    "ollb_hellaswag": -1,
+    "ollb_mmlu": -1,
+    "ollb_truthfulqa": -1
+  },
+  "codellama-34b": {
+    "category": "<40B",
+    "display_name": "Code LLaMA",
+    "thumb": "https://i.ibb.co/sPsWL9R/code-llama.png",
+    "thumb-mid": "https://i.ibb.co/RCv6nLV/code-llama-mid.png",
+    "thumb-tiny": "https://i.ibb.co/bFfcgtL/code-llama-tiny.png",
+    "parameters": "34",
+    "vram(full)": "65154",
+    "vram(8bit)": "36896",
+    "vram(4bit)": "21738",
+    "vram(gptq)": "N/A",
+    "hub(base)": "codellama/CodeLlama-34b-Instruct-hf",
+    "hub(ckpt)": "N/A",
+    "hub(gptq)": "N/A",
+    "hub(gptq_base)": "N/A",
+    "default_gen_config": "configs/response_configs/llama2.yaml",
+    "desc": "Code Llama is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 34 billion parameters. This is the repository for the base 7B version in the Hugging Face Transformers format. This model is designed for general code synthesis and understanding. Links to other models can be found in the index at the bottom.",
+    "example1": [],
+    "example2": [],
+    "example3": [],
+    "example4": [],
+    "ollb_average": -1,
+    "ollb_arc": -1,
+    "ollb_hellaswag": -1,
+    "ollb_mmlu": -1,
+    "ollb_truthfulqa": -1
   }
 }
\ No newline at end of file
diff --git a/utils.py b/utils.py
index 5b7347b..b408de6 100644
--- a/utils.py
+++ b/utils.py
@@ -184,7 +184,8 @@ def ctx(cls, context):
     def prompt(cls, pingpong, truncate_size):
         ping = pingpong.ping[:truncate_size]
         pong = "" if pingpong.pong is None else pingpong.pong[:truncate_size]
-        return f"""[INST] {ping} [/INST] {pong}
+        return f"""[INST] {ping} [/INST]
+{pong}
 """
 
 class LLaMA2ChatPPManager(PPManager):
@@ -514,7 +515,7 @@ def get_chat_manager(model_type):
         return GradioAlpacaChatPPManager()
     elif model_type == "alpaca-gpt4":
         return GradioAlpacaChatPPManager()
-    elif model_type == "nous-hermes":
+    elif model_type == "nous-hermes" or model_type == "nous-hermes2":
         return GradioAlpacaChatPPManager()
     elif model_type == "stablelm":
         return GradioStableLMChatPPManager()
@@ -560,7 +561,7 @@ def get_chat_manager(model_type):
         return GradioRedPajamaChatPPManager()
     elif model_type == "airoboros":
         return GradioVicunaChatPPManager()
-    elif model_type == "samantha-vicuna":
+    elif model_type == "samantha-vicuna" or model_type == "samantha2":
         return GradioVicunaChatPPManager()
     elif model_type == "lazarus":
         return GradioAlpacaChatPPManager()
@@ -572,11 +573,14 @@ def get_chat_manager(model_type):
         return GradioVicunaChatPPManager()
     elif model_type == "wizard-coder":
         return GradioAlpacaChatPPManager()
-    elif model_type == "orcamini":
+    elif model_type == "orcamini" or model_type == "orcamini2":
         return GradioOrcaMiniChatPPManager()
     elif model_type == "xgen":
         return GradioXGenChatPPManager()
-    elif model_type == "llama2":
+    elif model_type == "llama2" or \
+        model_type == "codellama" or \
+        model_type == "llama2-70b" or \
+        model_type == "codellama2-70b":
         return GradioLLaMA2ChatPPManager()
     elif model_type == "upstage-llama":
         return GradioAlpacaChatPPManager()
@@ -588,6 +592,8 @@ def get_chat_manager(model_type):
         return GradioPuffinChatPPManager()
     elif model_type == "platypus2":
         return GradioAlpacaChatPPManager()
+    elif model_type == "godzilla2":
+        return GradioAlpacaChatPPManager()
     else:
         return None
 
@@ -616,7 +622,10 @@ def get_global_context(model_type):
 It also MUST avoid giving false or misleading information, and it caveats when it isn’t entirely sure about the right answer.
 That said, the assistant is practical and really does its best, and doesn’t let caution get too much in the way of being useful.
 """
-    elif model_type == "llama2":
+    elif model_type == "llama2" or \
+        model_type == "codellama" or \
+        model_type == "llama2-70b" or \
+        model_type == "codellama2-70b":
         return """\
 You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe.  Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature.
 
@@ -626,7 +635,7 @@ def get_global_context(model_type):
     elif model_type == "xgen":
         return """A chat between a curious human and an artificial intelligence assistant.
 The assistant gives helpful, detailed, and polite answers to the human's questions."""
-    elif model_type == "orcamini":
+    elif model_type == "orcamini" or model_type == "orcamini2":
         return """You are an AI assistant that follows instruction extremely well. Help as much as you can.
 """
     elif model_type == "upstage-llama":
@@ -647,6 +656,15 @@ def get_global_context(model_type):
 It also MUST avoid giving false or misleading information, and it caveats when it isn’t entirely sure about the right answer.
 That said, the assistant is practical and really does its best, and doesn’t let caution get too much in the way of being useful.
 """
+    elif model_type == "godzilla2":
+        return """Below are a series of dialogues between human and an AI assistant.
+The AI tries to answer the given instruction as in response.
+The AI MUST not generate any text containing `### Response` or `### Instruction`.
+The AI MUST be helpful, polite, honest, sophisticated, emotionally aware, and humble-but-knowledgeable.
+The assistant MUST be happy to help with almost anything, and will do its best to understand exactly what is needed.
+It also MUST avoid giving false or misleading information, and it caveats when it isn’t entirely sure about the right answer.
+That said, the assistant is practical and really does its best, and doesn’t let caution get too much in the way of being useful.
+"""        
     elif model_type == "openllama":
         return """Below are a series of dialogues between human and an AI assistant.
 The AI tries to answer the given instruction as in response.
@@ -665,7 +683,7 @@ def get_global_context(model_type):
 It also MUST avoid giving false or misleading information, and it caveats when it isn’t entirely sure about the right answer.
 That said, the assistant is practical and really does its best, and doesn’t let caution get too much in the way of being useful.
 """
-    elif model_type == "nous-hermes":
+    elif model_type == "nous-hermes" or model_type == "nous-hermes2":
         return """Below are a series of dialogues between human and an AI assistant.
 The AI tries to answer the given instruction as in response.
 The AI MUST not generate any text containing `### Response` or `### Instruction`.
@@ -816,7 +834,7 @@ def get_global_context(model_type):
         return ""
     elif model_type == "puffin":
         return ""
-    elif model_type == "samantha-vicuna":
+    elif model_type == "samantha-vicuna" or model_type == "samantha2":
         return "You are Samantha, a sentient AI."
     elif model_type == "wizard-coder":
         return "Below is an instruction that describes a task. Write a response that appropriately completes the request."