added QA code/dataset

.env.example

@@ -1,2 +1,81 @@
-HF_TOKEN=
-MODEL_NAME=microsoft/Phi-3.5-mini-instruct
+LLM_MODEL_TYPE=huggingface
+# LLM_MODEL_TYPE=openai
+# LLM_MODEL_TYPE=hftgi
+# LLM_MODEL_TYPE=ollama
+# LLM_MODEL_TYPE=google
+# LLM_MODEL_TYPE=vllm
+
+HUGGINGFACE_AUTH_TOKEN=
+
+HFTGI_SERVER_URL=
+
+OPENAI_API_KEY=
+
+GOOGLE_API_KEY=
+
+# if unset, default to "gpt-3.5-turbo"
+OPENAI_MODEL_NAME=
+
+# GEMINI_MODEL_NAME=gemini-1.5-pro-latest
+
+# OLLAMA_MODEL_NAME=orca2:7b
+# OLLAMA_MODEL_NAME=mistral:7b
+# OLLAMA_MODEL_NAME=gemma:7b
+# OLLAMA_MODEL_NAME=llama2:7b
+OLLAMA_MODEL_NAME=llama3:8b
+
+OLLAMA_RP=1.15
+HF_RP=1.15
+
+LANGCHAIN_DEBUG=false
+BATCH_SIZE=1
+APPLY_CHAT_TEMPLATE_FOR_RAG=true
+
+# cpu, mps or cuda:0 - if unset, use whatever detected
+HF_EMBEDDINGS_DEVICE_TYPE=
+HF_PIPELINE_DEVICE_TYPE=
+
+# uncomment one of the below to load corresponding quantized model
+# LOAD_QUANTIZED_MODEL=4bit
+# LOAD_QUANTIZED_MODEL=8bit
+
+QA_WITH_RAG=true
+# QA_WITH_RAG=false
+
+RETRIEVER_TYPE=questions_file
+# RETRIEVER_TYPE=vectorstore
+
+QUESTIONS_FILE_PATH="./data/datasets/ms_macro.json"
+
+DISABLE_MODEL_PRELOADING=true
+CHAT_HISTORY_ENABLED=false
+SHOW_PARAM_SETTINGS=false
+SHARE_GRADIO_APP=false
+
+# if unset, default to "hkunlp/instructor-xl"
+HF_EMBEDDINGS_MODEL_NAME="hkunlp/instructor-large"
+
+# number of cpu cores - used to set n_threads for GPT4ALL & LlamaCpp models
+NUMBER_OF_CPU_CORES=
+
+USING_TORCH_BFLOAT16=true
+
+# HUGGINGFACE_MODEL_NAME_OR_PATH="databricks/dolly-v2-3b"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="databricks/dolly-v2-7b"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="databricks/dolly-v2-12b"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="TheBloke/wizardLM-7B-HF"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="TheBloke/vicuna-7B-1.1-HF"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="nomic-ai/gpt4all-j"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="nomic-ai/gpt4all-falcon"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="lmsys/fastchat-t5-3b-v1.0"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Llama-2-7b-chat-hf"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Llama-2-13b-chat-hf"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Llama-2-70b-chat-hf"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Meta-Llama-3-8B-Instruct"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Meta-Llama-3-70B-Instruct"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="microsoft/Orca-2-7b"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="microsoft/Orca-2-13b"
+HUGGINGFACE_MODEL_NAME_OR_PATH="google/gemma-1.1-2b-it"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="google/gemma-1.1-7b-it"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="microsoft/Phi-3-mini-128k-instruct"
+# HUGGINGFACE_MODEL_NAME_OR_PATH="mistralai/Mistral-7B-Instruct-v0.2"

app.py

@@ -13,7 +13,7 @@ if len(found_dotenv) == 0:
 print(f"loading env vars from: {found_dotenv}")
 load_dotenv(found_dotenv, override=False)
 
-path = os.path.dirname(found_dotenv)
+path = os.path.dirname(found_dotenv) + "/src"
 print(f"Adding {path} to sys.path")
 sys.path.append(path)
 
@@ -24,7 +24,7 @@ hf_token = os.getenv("HF_TOKEN")
 
 login(token=hf_token, add_to_git_credential=True)
 
-questions_file_path = os.getenv("QUESTIONS_FILE_PATH") or "./ms_macro.json"
+questions_file_path = os.getenv("QUESTIONS_FILE_PATH") or "./data/datasets/ms_macro.json"
 
 questions = json.loads(open(questions_file_path).read())
 examples = [[question["question"].strip()] for question in questions]
@@ -46,8 +46,6 @@ def chat(
     history: list[tuple[str, str]],
     system_message,
     temperature=0,
-    frequency_penalty=0,
-    presence_penalty=0,
     max_tokens=256,
     top_p=0.95,
 ):
@@ -76,8 +74,6 @@ def chat(
         max_tokens=max_tokens,
         stream=True,
         temperature=temperature,
-        frequency_penalty=None,  # frequency_penalty,
-        presence_penalty=None,  # presence_penalty,
         top_p=top_p,
         seed=42,
     ):
@@ -135,22 +131,6 @@ demo = gr.ChatInterface(
         gr.Slider(
             minimum=0, maximum=2, step=0.1, value=0, label="Temperature", render=False
         ),
-        gr.Slider(
-            minimum=-2,
-            maximum=2,
-            step=0.1,
-            value=0,
-            label="Frequency Penalty",
-            render=False,
-        ),
-        gr.Slider(
-            minimum=-2,
-            maximum=2,
-            step=0.1,
-            value=0,
-            label="Presence Penalty",
-            render=False,
-        ),
         gr.Slider(
             minimum=128,
             maximum=4096,

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa33b726196f9c503102e158c7333a50f86635a0bc5a8b7b1cfdb8e435d65255
+size 4143

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.00.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e74664d615b6f9ff601acad90ec8baad890aed8736629451f6f8685001e85ffb
+size 517847

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.02.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:557848e5f0f54011eaeedd9c61dc836d2e4d313c7128a552a552a46a9ae2c906
+size 512769

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.04.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17ecc5d4cd300069c22b7f9a3e1b4ae10155268b9d7d4baee0a85c8a936cb139
+size 512204

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.06.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e0f75e74d37cce10fe5be81f722e99f94582fe69298d7bbe8407343c9d87b89
+size 515412

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.08.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89952f9332501697e061e008b5b5e4193bb2f1ec620a9d9fa6c5dc2195e4f359
+size 525035

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.10.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:141cb33fc2cd887bda0cdad3870792921e06c08edb81b53819d8df8f990c5979
+size 519175

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.12.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e38ad96f0f641102dc5053deb01fe044ae7738c8bb7d728804cc94004f6620e2
+size 519424

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.14.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b0f5fd8b04c198a3e4cd5c8deac367092438ecc20f533f9fab9d357b1d3edd1
+size 524587

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.16.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:567ed6c25205bb8a3a5468b618edba4fac3620e6621b24fa4e949749d31daf7b
+size 524120

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.18.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3af977ebea2d2a202ab481b8604edc8034fe1d34f1e2c1443a06b9b18bd240a9
+size 527908

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.20.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1ed7a7addff35dcaf5522fa9df06eba02acbc2194d306745224d0a47cc35a21
+size 526303

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.22.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34376c053817713b90bf6cd87c38043527c597090c9cd2688316a2e57453764b
+size 530616

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.24.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fa075c2c32a4b1ebc1d44b85ab0c3a3409eb6b13030c54ab83d10303ed856e9
+size 533097

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.26.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f15060426386580723b98461f9b13524c06549ef4e5934b8a5ad9b527c75c67
+size 527490

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.28.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd95f711c790ca57ff9cebf715bd83979d9695563277839225736ced49aa3d7c
+size 541483

data/results/Llama-2-13b-chat-hf(Non-RAG)_mm_rpp_1.30.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49d17bc145cfb3461e95193db4664a6fa0ac8a89837aedde95f20748b84999ed
+size 537067

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52f034c39872e94a5daa4fa0edbdfb9eac6cea77abbfc205a6fc68de14c89f05
+size 4719

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.00.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d781812ba2430d6fd9d66994fe2684db95681d4a32717095e6379f5cb7010a61
+size 912531

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.02.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c99bfb50ae2788fe86620c8b11a208b561d1db9c041baa80aa877215b18060b5
+size 911612

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.04.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69f074123658fabb2c23c00bc17a3c899f78f807ff61d225fdd3b64b8b62f56c
+size 905187

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.06.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1eea92d5985d45e9ebfcdae66103fe6230980b941499e954b524b8f9ae16d9bd
+size 910859

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.08.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a929daf31182f5d359bcef12df91e51fd0101e09831383dd4e784113a175379a
+size 910911

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.10.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f25047130eeb8575c581b33b372bfe48e365a95638d96a10107ad3d68456e8f0
+size 906530

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.12.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1abc3770c959ae2414c4c45540b156893f2aa55a5d01f85647f619052ca88a64
+size 895938

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.14.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05f8fa1d15915894634921d0e6f50e3007de964ab736dc6d86106635ce264493
+size 899020

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.16.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b13dfdaa7e8aa4eeea2563565f900161b4dfc79fb32e780358601826d646025c
+size 903763

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.18.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:778cfb3c73c2f4ef07cb354f95315152f92812f6f2a5cd545ef942916cdd8695
+size 902295

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.20.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:daffa3297c39e7099c2f4fed1a3128bbf0890405fbb2a288474eae2a57dee0c6
+size 902764

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.22.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e2be7b610927f1e4e5a68f3000fed30b9d8345344b63d84162af654b397b767
+size 907375

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.24.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30372004528146170f8413af65be3c88da94d3e4df0e580da1c329f48229d703
+size 911356

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.26.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd3b9acf88c9c94b17ddb753122345e18defaed5d0e0eced30eab36b6767142c
+size 913735

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.28.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a30a9742a8790c5f76fecf78ebd6543be33c63e2015a925bac649e7fde856277
+size 913015

data/results/Llama-2-13b-chat-hf(Non-RAG)_wd_rpp_1.30.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30fe001a0fb83fc1d36f0eeca3512492745de5d1815af69238a109f1ab624547
+size 929776

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e28af86ab18a1f43ddf76c90aab009803241d602dec46bab30c697c2ff42847a
+size 4187

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.00.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a0d5338273f84ef7a28e16e6b86b90d92794acfaa72dc0469d82e0ba51ac8b7
+size 399758

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.02.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7c3ab09f1b691ff6dc29c985bc5086fde4e4c208ac4365fda7699e6a25b15bc
+size 398942

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.04.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68b16d30ddcdc6fd74f8dae37c7d1ed939f6127b14bf20a03bf62f7b2e7cd28b
+size 401963

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.06.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:480442e76bc506a0b59d82d97821b46c4d826fd0e06d69b8452b8f3729acd387
+size 398890

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.08.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57bd6f4c295d161f4fa3d28e9ca314a6839a6685bf604cc255f17cdbbabe675b
+size 404554

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.10.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4df87cd1556292fb63fc018290aad1c2c2db379a3d17c073373c6b3622aa114f
+size 398221

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.12.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22ce81e92dc0ca4d1959c811730b04ccfc64e7fe10446b3ed6855b47c283c065
+size 398798

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.14.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:516afe0b4689104160a09e1d0c1b75718bd3c1a6b62781baecfe8a64bea9b69c
+size 404891

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.16.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b475278a5b6a252f7192d54801b67b1a6fa44998ab85866a334dd8d26eaac20
+size 399407

data/results/Llama-2-13b-chat-hf(RAG - Chat Template)_mm_rpp_1.18.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41bad8662eca9dcbc8c0b898aec8b6a85c484c7a7141b43557bcca369519b4fc
+size 398046