update

Reviewed-on: #3

Manifest.toml

@@ -1,8 +1,8 @@
 # This file is machine-generated - editing it directly is not advised
 
-julia_version = "1.11.2"
+julia_version = "1.11.4"
 manifest_format = "2.0"
-project_hash = "b483014657ef9f0fde60d7258585b291d6f0eeca"
+project_hash = "cb7f3c57318e927e8ac4dc2dea9acdcace566ed1"
 
 [[deps.AliasTables]]
 deps = ["PtrArrays", "Random"]
@@ -120,9 +120,9 @@ version = "1.11.0"
 
 [[deps.Distributions]]
 deps = ["AliasTables", "FillArrays", "LinearAlgebra", "PDMats", "Printf", "QuadGK", "Random", "SpecialFunctions", "Statistics", "StatsAPI", "StatsBase", "StatsFuns"]
-git-tree-sha1 = "3101c32aab536e7a27b1763c0797dba151b899ad"
+git-tree-sha1 = "0b4190661e8a4e51a842070e7dd4fae440ddb7f4"
 uuid = "31c24e10-a181-5473-b8eb-7969acd0382f"
-version = "0.25.113"
+version = "0.25.118"
 
     [deps.Distributions.extensions]
     DistributionsChainRulesCoreExt = "ChainRulesCore"
@@ -158,9 +158,9 @@ version = "0.1.10"
 
 [[deps.FileIO]]
 deps = ["Pkg", "Requires", "UUIDs"]
-git-tree-sha1 = "2dd20384bf8c6d411b5c7370865b1e9b26cb2ea3"
+git-tree-sha1 = "b66970a70db13f45b7e57fbda1736e1cf72174ea"
 uuid = "5789e2e9-d7fb-5bc7-8068-2c6fae9b9549"
-version = "1.16.6"
+version = "1.17.0"
 weakdeps = ["HTTP"]
 
     [deps.FileIO.extensions]
@@ -168,9 +168,9 @@ weakdeps = ["HTTP"]
 
 [[deps.FilePathsBase]]
 deps = ["Compat", "Dates"]
-git-tree-sha1 = "7878ff7172a8e6beedd1dea14bd27c3c6340d361"
+git-tree-sha1 = "3bab2c5aa25e7840a4b065805c0cdfc01f3068d2"
 uuid = "48062228-2e41-5def-b9a4-89aafe57970f"
-version = "0.9.22"
+version = "0.9.24"
 weakdeps = ["Mmap", "Test"]
 
     [deps.FilePathsBase.extensions]
@@ -200,11 +200,9 @@ version = "1.11.0"
 
 [[deps.GeneralUtils]]
 deps = ["CSV", "DataFrames", "DataStructures", "Dates", "Distributions", "JSON3", "MQTTClient", "PrettyPrinting", "Random", "SHA", "UUIDs"]
-git-tree-sha1 = "978d9a5c3fc30205dd72d4a2a2ed4fa85ebee5cf"
+path = "/appfolder/app/dev/GeneralUtils"
-repo-rev = "main"
-repo-url = "https://git.yiem.cc/ton/GeneralUtils"
 uuid = "c6c72f09-b708-4ac8-ac7c-2084d70108fe"
-version = "0.1.0"
+version = "0.2.3"
 
 [[deps.HTTP]]
 deps = ["Base64", "CodecZlib", "ConcurrentUtilities", "Dates", "ExceptionUnwrapping", "Logging", "LoggingExtras", "MbedTLS", "NetworkOptions", "OpenSSL", "PrecompileTools", "Random", "SimpleBufferStream", "Sockets", "URIs", "UUIDs"]
@@ -214,9 +212,9 @@ version = "1.10.13"
 
 [[deps.HypergeometricFunctions]]
 deps = ["LinearAlgebra", "OpenLibm_jll", "SpecialFunctions"]
-git-tree-sha1 = "b1c2585431c382e3fe5805874bda6aea90a95de9"
+git-tree-sha1 = "68c173f4f449de5b438ee67ed0c9c748dc31a2ec"
 uuid = "34004b35-14d8-5ef3-9330-4cdb6864b03a"
-version = "0.3.25"
+version = "0.3.28"
 
 [[deps.ICU_jll]]
 deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
@@ -260,9 +258,9 @@ uuid = "41ab1584-1d38-5bbf-9106-f11c6c58b48f"
 version = "1.3.0"
 
 [[deps.IrrationalConstants]]
-git-tree-sha1 = "630b497eafcc20001bba38a4651b327dcfc491d2"
+git-tree-sha1 = "e2222959fbc6c19554dc15174c81bf7bf3aa691c"
 uuid = "92d709cd-6900-40b7-9082-c6be49f344b6"
-version = "0.2.2"
+version = "0.2.4"
 
 [[deps.IterTools]]
 git-tree-sha1 = "42d5f897009e7ff2cf88db414a389e5ed1bdd023"
@@ -305,12 +303,10 @@ uuid = "b39eb1a6-c29a-53d7-8c32-632cd16f18da"
 version = "1.19.3+0"
 
 [[deps.LLMMCTS]]
-deps = ["GeneralUtils", "JSON3"]
+deps = ["GeneralUtils", "JSON3", "PrettyPrinting"]
-git-tree-sha1 = "d8c653b8fafbd3757b7332985efaf1fdb8b6fe97"
+path = "/appfolder/app/dev/LLMMCTS"
-repo-rev = "main"
-repo-url = "https://git.yiem.cc/ton/LLMMCTS"
 uuid = "d76c5a4d-449e-4835-8cc4-dd86ec44f241"
-version = "0.1.2"
+version = "0.1.4"
 
 [[deps.LaTeXStrings]]
 git-tree-sha1 = "dda21b8cbd6a6c40d9d02a73230f9d70fed6918c"
@@ -370,9 +366,9 @@ version = "1.11.0"
 
 [[deps.LogExpFunctions]]
 deps = ["DocStringExtensions", "IrrationalConstants", "LinearAlgebra"]
-git-tree-sha1 = "a2d09619db4e765091ee5c6ffe8872849de0feea"
+git-tree-sha1 = "13ca9e2586b89836fd20cccf56e57e2b9ae7f38f"
 uuid = "2ab3a3ac-af41-5b50-aa03-7779005ae688"
-version = "0.3.28"
+version = "0.3.29"
 
     [deps.LogExpFunctions.extensions]
     LogExpFunctionsChainRulesCoreExt = "ChainRulesCore"
@@ -475,7 +471,7 @@ version = "0.3.27+1"
 [[deps.OpenLibm_jll]]
 deps = ["Artifacts", "Libdl"]
 uuid = "05823500-19ac-5b8b-9628-191a04bc5112"
-version = "0.8.1+2"
+version = "0.8.1+4"
 
 [[deps.OpenSSL]]
 deps = ["BitFlags", "Dates", "MozillaCACerts_jll", "OpenSSL_jll", "Sockets"]
@@ -493,7 +489,7 @@ version = "3.0.15+1"
 deps = ["Artifacts", "CompilerSupportLibraries_jll", "JLLWrappers", "Libdl", "Pkg"]
 git-tree-sha1 = "13652491f6856acfd2db29360e1bbcd4565d04f1"
 uuid = "efe28fd5-8261-553b-a9e1-b2916fc3738e"
-version = "0.5.5+0"
+version = "0.5.5+2"
 
 [[deps.OrderedCollections]]
 git-tree-sha1 = "12f1439c4f986bb868acda6ea33ebc78e19b95ad"
@@ -502,9 +498,9 @@ version = "1.7.0"
 
 [[deps.PDMats]]
 deps = ["LinearAlgebra", "SparseArrays", "SuiteSparse"]
-git-tree-sha1 = "949347156c25054de2db3b166c52ac4728cbad65"
+git-tree-sha1 = "48566789a6d5f6492688279e22445002d171cf76"
 uuid = "90014a1f-27ba-587c-ab20-58faa44d9150"
-version = "0.11.31"
+version = "0.11.33"
 
 [[deps.Parsers]]
 deps = ["Dates", "PrecompileTools", "UUIDs"]
@@ -556,15 +552,15 @@ uuid = "de0858da-6303-5e67-8744-51eddeeeb8d7"
 version = "1.11.0"
 
 [[deps.PtrArrays]]
-git-tree-sha1 = "77a42d78b6a92df47ab37e177b2deac405e1c88f"
+git-tree-sha1 = "1d36ef11a9aaf1e8b74dacc6a731dd1de8fd493d"
 uuid = "43287f4e-b6f4-7ad1-bb20-aadabca52c3d"
-version = "1.2.1"
+version = "1.3.0"
 
 [[deps.QuadGK]]
 deps = ["DataStructures", "LinearAlgebra"]
-git-tree-sha1 = "cda3b045cf9ef07a08ad46731f5a3165e56cf3da"
+git-tree-sha1 = "9da16da70037ba9d701192e27befedefb91ec284"
 uuid = "1fd47b50-473d-5c70-9696-f719f8f3bcdc"
-version = "2.11.1"
+version = "2.11.2"
 
     [deps.QuadGK.extensions]
     QuadGKEnzymeExt = "Enzyme"
@@ -623,11 +619,9 @@ version = "0.7.0"
 
 [[deps.SQLLLM]]
 deps = ["CSV", "DataFrames", "DataStructures", "Dates", "FileIO", "GeneralUtils", "HTTP", "JSON3", "LLMMCTS", "LibPQ", "PrettyPrinting", "Random", "Revise", "StatsBase", "Tables", "URIs", "UUIDs"]
-git-tree-sha1 = "45e660e44de0950a5e5f92d467298d8b768b6023"
+path = "/appfolder/app/dev/SQLLLM"
-repo-rev = "main"
-repo-url = "https://git.yiem.cc/ton/SQLLLM"
 uuid = "2ebc79c7-cc10-4a3a-9665-d2e1d61e63d3"
-version = "0.2.0"
+version = "0.2.4"
 
 [[deps.SQLStrings]]
 git-tree-sha1 = "55de0530689832b1d3d43491ee6b67bd54d3323c"
@@ -672,9 +666,9 @@ version = "1.11.0"
 
 [[deps.SpecialFunctions]]
 deps = ["IrrationalConstants", "LogExpFunctions", "OpenLibm_jll", "OpenSpecFun_jll"]
-git-tree-sha1 = "2f5d4697f21388cbe1ff299430dd169ef97d7e14"
+git-tree-sha1 = "64cca0c26b4f31ba18f13f6c12af7c85f478cfde"
 uuid = "276daf66-3868-5448-9aa4-cd146d93841b"
-version = "2.4.0"
+version = "2.5.0"
 
     [deps.SpecialFunctions.extensions]
     SpecialFunctionsChainRulesCoreExt = "ChainRulesCore"
@@ -699,16 +693,16 @@ uuid = "82ae8749-77ed-4fe6-ae5f-f523153014b0"
 version = "1.7.0"
 
 [[deps.StatsBase]]
-deps = ["DataAPI", "DataStructures", "LinearAlgebra", "LogExpFunctions", "Missings", "Printf", "Random", "SortingAlgorithms", "SparseArrays", "Statistics", "StatsAPI"]
+deps = ["AliasTables", "DataAPI", "DataStructures", "LinearAlgebra", "LogExpFunctions", "Missings", "Printf", "Random", "SortingAlgorithms", "SparseArrays", "Statistics", "StatsAPI"]
-git-tree-sha1 = "5cf7606d6cef84b543b483848d4ae08ad9832b21"
+git-tree-sha1 = "29321314c920c26684834965ec2ce0dacc9cf8e5"
 uuid = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
-version = "0.34.3"
+version = "0.34.4"
 
 [[deps.StatsFuns]]
 deps = ["HypergeometricFunctions", "IrrationalConstants", "LogExpFunctions", "Reexport", "Rmath", "SpecialFunctions"]
-git-tree-sha1 = "b423576adc27097764a90e163157bcfc9acf0f46"
+git-tree-sha1 = "35b09e80be285516e52c9054792c884b9216ae3c"
 uuid = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
-version = "1.3.2"
+version = "1.4.0"
 
     [deps.StatsFuns.extensions]
     StatsFunsChainRulesCoreExt = "ChainRulesCore"

Project.toml

@@ -1,9 +1,10 @@
 name = "YiemAgent"
 uuid = "e012c34b-7f78-48e0-971c-7abb83b6f0a2"
 authors = ["narawat lamaiin <narawat@outlook.com>"]
-version = "0.1.4"
+version = "0.2.0"
 
 [deps]
+CSV = "336ed68f-0bac-5ca0-87d4-7b16caf5d00b"
 DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
 DataStructures = "864edb3b-99cc-5e75-8d2d-829cb0a9cfe8"
 Dates = "ade2ca70-3891-5945-98fb-dc099432e06a"
@@ -21,7 +22,5 @@ URIs = "5c2747f8-b7ea-4ff2-ba2e-563bfd36b1d4"
 UUIDs = "cf7118a7-6976-5b1a-9a39-7adc72f591a4"
 
 [compat]
+CSV = "0.10.15"
 DataFrames = "1.7.0"
-GeneralUtils = "0.1, 0.2"
-LLMMCTS = "0.1.2"
-SQLLLM = "0.2.0"

fast_inference_guideline.txt

@@ -0,0 +1,72 @@
+To make **LLM-driven inference** fast while maintaining its dynamic capabilities, there are a few practices or approaches to avoid, as they could lead to performance bottlenecks or inefficiencies. Here's what *not* to do:
+
+---
+
+### **1. Avoid Using Overly Large Models for Every Query**
+While larger LLMs like GPT-4 provide high accuracy and nuanced responses, they may slow down real-time processing due to their computational complexity. Instead:
+- Use distilled or smaller models (e.g., GPT-3.5 Turbo or fine-tuned versions) for faster inference without compromising much on quality.
+
+---
+
+### **2. Avoid Excessive Entity Preprocessing**
+Don’t rely on overly complicated preprocessing steps (like advanced NER models or regex-heavy pipelines) to extract entities from the query before invoking the LLM. This could add latency. Instead:
+- Design efficient prompts that allow the LLM to extract entities and generate responses simultaneously.
+
+---
+
+### **3. Avoid Asking the LLM Multiple Separate Questions**
+Running the LLM for multiple subtasks—for example, entity extraction first and response generation second—can significantly slow down the pipeline. Instead:
+- Create prompts that combine tasks into one pass, e.g., *"Identify the city name and generate a weather response for this query: 'What's the weather in London?'"*.
+
+---
+
+### **4. Don’t Overload the LLM with Context History**
+Excessively lengthy conversation history or irrelevant context in your prompts can slow down inference times. Instead:
+- Provide only the relevant context for each query, trimming unnecessary parts of the conversation.
+
+---
+
+### **5. Avoid Real-Time Dependence on External APIs**
+Using external APIs to fetch supplementary data (e.g., weather details or location info) during every query can introduce latency. Instead:
+- Pre-fetch API data asynchronously and use the LLM to integrate it dynamically into responses.
+
+---
+
+### **6. Avoid Running LLM on Underpowered Hardware**
+Running inference on CPUs or low-spec GPUs will result in slower response times. Instead:
+- Deploy the LLM on optimized infrastructure (e.g., high-performance GPUs like NVIDIA A100 or cloud platforms like Azure AI) to reduce latency.
+
+---
+
+### **7. Skip Lengthy Generative Prompts**
+Avoid prompts that encourage the LLM to produce overly detailed or verbose responses, as these take longer to process. Instead:
+- Use concise prompts that focus on generating actionable or succinct answers.
+
+---
+
+### **8. Don’t Ignore Optimization Techniques**
+Failing to optimize your LLM setup can drastically impact performance. For example:
+- Avoid skipping techniques like model quantization (reducing numerical precision to speed up inference) or distillation (training smaller models).
+
+---
+
+### **9. Don’t Neglect Response Caching**
+While you may not want a full caching system to avoid sunk costs, dismissing lightweight caching entirely can impact speed. Instead:
+- Use temporary session-based caching for very frequent queries, without committing to a full-fledged cache infrastructure.
+
+---
+
+### **10. Avoid One-Size-Fits-All Solutions**
+Applying the same LLM inference method to all queries—whether simple or complex—will waste processing resources. Instead:
+- Route basic queries to faster, specialized models and use the LLM for nuanced or multi-step queries only.
+
+---
+
+### Summary: Focus on Efficient Design
+By avoiding these pitfalls, you can ensure that LLM-driven inference remains fast and responsive:
+- Optimize prompts.
+- Use smaller models for simpler queries.
+- Run the LLM on high-performance hardware.
+- Trim unnecessary preprocessing or contextual steps.
+
+Would you like me to help refine a prompt or suggest specific tools to complement your implementation? Let me know!

src/llmfunction.jl

@@ -296,13 +296,14 @@ function checkinventory(a::T1, input::T2
   wineattributes_2 = extractWineAttributes_2(a, input)
 
   _inventoryquery = "retailer name: $(a.retailername), $wineattributes_1, $wineattributes_2"
-  inventoryquery = "Retrieves winery, wine_name, vintage, region, country, wine_type, grape, serving_temperature, sweetness, intensity, tannin, acidity, tasting_notes, price and currency of wines that match the following criteria - {$_inventoryquery}"
+  inventoryquery = "Retrieves winery, wine_name, wine_id, vintage, region, country, wine_type, grape, serving_temperature, sweetness, intensity, tannin, acidity, tasting_notes, price and currency of wines that match the following criteria - {$_inventoryquery}"
   println("\ncheckinventory input: $inventoryquery ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
   # add suppport for similarSQLVectorDB
   textresult, rawresponse = SQLLLM.query(inventoryquery, a.func[:executeSQL], 
-                        a.func[:text2textInstructLLM],
+                        a.func[:text2textInstructLLM];
                         insertSQLVectorDB=a.func[:insertSQLVectorDB], 
-                        similarSQLVectorDB=a.func[:similarSQLVectorDB])
+                        similarSQLVectorDB=a.func[:similarSQLVectorDB],
+                        llmFormatName="qwen3")
 
   println("\ncheckinventory result ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
   println(textresult)
@@ -330,7 +331,8 @@ julia>
 
 # Signature
 """
-function extractWineAttributes_1(a::T1, input::T2)::String where {T1<:agent, T2<:AbstractString}
+function extractWineAttributes_1(a::T1, input::T2; maxattempt=10
+  )::String where {T1<:agent, T2<:AbstractString}
 
   systemmsg =
   """
@@ -358,45 +360,50 @@ function extractWineAttributes_1(a::T1, input::T2)::String where {T1<:agent, T2<
     Occasion: the occasion the user is having the wine for
     Food_to_be_paired_with_wine: food that the user will be served with the wine such as poultry, fish, steak, etc
 
-  You should only respond in format as described below:
+    You should only respond in format as described below:
-  Thought: ...
+    Thought: ...
-  Wine_name: ...
+    Wine_name: ...
-  Winery: ...
+    Winery: ...
-  Vintage: ...
+    Vintage: ...
-  Region: ...
+    Region: ...
-  Country: ...
+    Country: ...
-  Wine_type:
+    Wine_type:
-  Grape_varietal: ...
+    Grape_varietal: ...
-  Tasting_notes: ...
+    Tasting_notes: ...
-  Wine_price: ...
+    Wine_price: ...
-  Occasion: ...
+    Occasion: ...
-  Food_to_be_paired_with_wine: ...
+    Food_to_be_paired_with_wine: ...
-  
+    
-  Here are some example:
+    Here are some example:
-  User's query: red, Chenin Blanc, Riesling, 20 USD
+    User's query: red, Chenin Blanc, Riesling, 20 USD
-  {"reasoning": ..., "winery": "NA", "wine_name": "NA", "vintage": "NA", "region": "NA", "country": "NA", "wine_type": "red, white", "grape_varietal": "Chenin Blanc, Riesling", "tasting_notes": "NA", "wine_price": "0-20", "occasion": "NA", "food_to_be_paired_with_wine": "NA"}
+    {"reasoning": ..., "winery": "NA", "wine_name": "NA", "vintage": "NA", "region": "NA", "country": "NA", "wine_type": "red, white", "grape_varietal": "Chenin Blanc, Riesling", "tasting_notes": "NA", "wine_price": "0-20", "occasion": "NA", "food_to_be_paired_with_wine": "NA"}
 
-  User's query: Domaine du Collier Saumur Blanc 2019, France, white, Merlot
+    User's query: Domaine du Collier Saumur Blanc 2019, France, white, Merlot
-  {"reasoning": ..., "winery": "Domaine du Collier", "wine_name": "Saumur Blanc", "vintage": "2019", "region": "Saumur", "country": "France", "wine_type": "white", "grape_varietal": "Merlot", "tasting_notes": "NA", "wine_price": "NA", "occasion": "NA", "food_to_be_paired_with_wine": "NA"}
+    {"reasoning": ..., "winery": "Domaine du Collier", "wine_name": "Saumur Blanc", "vintage": "2019", "region": "Saumur", "country": "France", "wine_type": "white", "grape_varietal": "Merlot", "tasting_notes": "NA", "wine_price": "NA", "occasion": "NA", "food_to_be_paired_with_wine": "NA"}
 
-  Let's begin!
+    Let's begin!
   """
 
   header = ["Thought:", "Wine_name:", "Winery:", "Vintage:", "Region:", "Country:", "Wine_type:", "Grape_varietal:", "Tasting_notes:", "Wine_price:", "Occasion:", "Food_to_be_paired_with_wine:"]
   dictkey = ["thought", "wine_name", "winery", "vintage", "region", "country", "wine_type", "grape_varietal", "tasting_notes", "wine_price", "occasion", "food_to_be_paired_with_wine"]
-  errornote = ""
+  errornote = "N/A"
   
-  for attempt in 1:10
+  llmkwargs=Dict(
-    #[WORKING] I should add generatequestion()
+    :num_ctx => 32768,
+    :temperature => 0.5,
+  )
+
+  for attempt in 1:maxattempt
+    #[PENDING] I should add generatequestion()
 
     if attempt > 1
-      println("\nYiemAgent extractWineAttributes_1() attempt $attempt/10 ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
+      println("\nYiemAgent extractWineAttributes_1() attempt $attempt/$maxattempt ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
     end
 
     usermsg =
     """
     User's query: $input
-    $errornote
+    P.S. $errornote
     """
 
     _prompt = 
@@ -406,29 +413,33 @@ function extractWineAttributes_1(a::T1, input::T2)::String where {T1<:agent, T2<
     ]
 
     # put in model format
-    prompt = GeneralUtils.formatLLMtext(_prompt; formatname="qwen")
+    prompt = GeneralUtils.formatLLMtext(_prompt, "granite3")
-    response = a.func[:text2textInstructLLM](prompt)
+    response = a.func[:text2textInstructLLM](prompt; 
+      modelsize="medium", llmkwargs=llmkwargs, senderId=a.id)
     response = GeneralUtils.remove_french_accents(response)
+    response = GeneralUtils.deFormatLLMtext(response, "granite3")
 
     # check wheter all attributes are in the response
     checkFlag = false
     for word in header
       if !occursin(word, response)
-        errornote = "$word attribute is missing in previous attempts"
+        errornote = "In your previous attempts, the $word attribute is missing. Please try again."
-        println("Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
+        println("\nYiemAgent extractWineAttributes_1() Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
         checkFlag = true
         break
       end
     end
     checkFlag == true ? continue : nothing
     
-    # check whether response has all header
+    # check whether response has all answer's key points
     detected_kw = GeneralUtils.detect_keyword(header, response)
     if 0 ∈ values(detected_kw)
-      errornote = "\nYiemAgent extractWineAttributes_1() response does not have all header"
+      errornote = "In your previous attempts, the response does not have all answer's key points"
+      println("\nYiemAgent extractWineAttributes_1() Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
       continue
     elseif sum(values(detected_kw)) > length(header)
-      errornote = "\nYiemAgent extractWineAttributes_1() response has duplicated header"
+      errornote = "In your previous attempts, the response has duplicated answer's key points"
+      println("\nYiemAgent extractWineAttributes_1() Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
       continue
     end
     responsedict = GeneralUtils.textToDict(response, header; 
@@ -452,8 +463,8 @@ function extractWineAttributes_1(a::T1, input::T2)::String where {T1<:agent, T2<
           if responsedict[:wine_price] != "NA" 
             # check whether wine_price is in ranged number
             if !occursin('-', responsedict[:wine_price]) 
-              errornote = "wine_price must be a range number"
+              errornote = "In your previous attempt, the 'wine_price' was not set to a ranged number. Please adjust it accordingly."
-              println("ERROR YiemAgent extractWineAttributes_1() $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
+              println("\nERROR YiemAgent extractWineAttributes_1() $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
               checkFlag = true
               break
             end
@@ -467,8 +478,8 @@ function extractWineAttributes_1(a::T1, input::T2)::String where {T1<:agent, T2<
             end
             # price range like 100-100 is not good
             if minprice == maxprice
-              errornote = "wine_price with minimum equals to maximum is not valid"
+              errornote = "In your previous attempt, you inputted 'wine_price' with a 'minimum' value equaling the 'maximum', which is not valid."
-              println("ERROR YiemAgent extractWineAttributes_1() $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
+              println("\nERROR YiemAgent extractWineAttributes_1() $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
               checkFlag = true
               break
             end
@@ -568,14 +579,12 @@ function extractWineAttributes_2(a::T1, input::T2)::String where {T1<:agent, T2<
   The preference form requires the following information:
   sweetness, acidity, tannin, intensity
 
-  <You must follow the following guidelines>
+  You must follow the following guidelines:
     1) If specific information required in the preference form is not available in the query or there isn't any, mark with 'NA' to indicate this.
       Additionally, words like 'any' or 'unlimited' mean no information is available.
     2) Use the conversion table to convert the descriptive word level of sweetness, intensity, tannin, and acidity into a corresponding integer.
     3) Do not generate other comments.
-  </You must follow the following guidelines>
+  You should then respond to the user with:
-
-  <You should then respond to the user with>
     Sweetness_keyword: The exact keywords in the user's query describing the sweetness level of the wine.
     Sweetness: ( S ), where ( S ) represents integers indicating the range of sweetness levels. Example: 1-2
     Acidity_keyword: The exact keywords in the user's query describing the acidity level of the wine.
@@ -584,9 +593,7 @@ function extractWineAttributes_2(a::T1, input::T2)::String where {T1<:agent, T2<
     Tannin: ( T ), where ( T ) represents integers indicating the range of tannin level. Example: 1-3
     Intensity_keyword: The exact keywords in the user's query describing the intensity level of the wine.
     Intensity: ( I ), where ( I ) represents integers indicating the range of intensity level. Example: 2-4
-  </You should then respond to the user with>
+  You should only respond in format as described below:
-
-  <You should only respond in format as described below>
     Sweetness_keyword: ...
     Sweetness: ...
     Acidity_keyword: ...
@@ -595,9 +602,8 @@ function extractWineAttributes_2(a::T1, input::T2)::String where {T1<:agent, T2<
     Tannin: ...
     Intensity_keyword: ...
     Intensity: ...
-  </You should only respond in format as described below>
+    
-
+  Here are some examples:
-  <Here are some examples>
     User's query: I want a wine with a medium-bodied, low acidity, medium tannin.
     Sweetness_keyword: NA
     Sweetness: NA
@@ -617,7 +623,6 @@ function extractWineAttributes_2(a::T1, input::T2)::String where {T1<:agent, T2<
     Tannin: NA
     Intensity_keyword: NA
     Intensity: NA
-  </Here are some examples>
 
   Let's begin!
   """
@@ -630,7 +635,7 @@ function extractWineAttributes_2(a::T1, input::T2)::String where {T1<:agent, T2<
     """
     $conversiontable
     User's query: $input
-    $errornote
+    P.S. $errornote
     """
 
     _prompt = 
@@ -640,17 +645,20 @@ function extractWineAttributes_2(a::T1, input::T2)::String where {T1<:agent, T2<
     ]
 
     # put in model format
-    prompt = GeneralUtils.formatLLMtext(_prompt; formatname="qwen")
+    prompt = GeneralUtils.formatLLMtext(_prompt, "granite3")
 
     response = a.func[:text2textInstructLLM](prompt)
+    response = GeneralUtils.deFormatLLMtext(response, "granite3")
     
-    # check whether response has all header
+    # check whether response has all answer's key points
     detected_kw = GeneralUtils.detect_keyword(header, response)
     if 0 ∈ values(detected_kw)
-      errornote = "\nYiemAgent extractWineAttributes_2() response does not have all header"
+      errornote = "In your previous attempt does not have all answer's key points"
+      println("\nERROR YiemAgent extractWineAttributes_2() Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
       continue
     elseif sum(values(detected_kw)) > length(header)
-      errornote = "\nYiemAgent extractWineAttributes_2() response has duplicated header"
+      errornote = "In your previous attempt has duplicated answer's key points"
+      println("\nERROR YiemAgent extractWineAttributes_2() Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
       continue
     end
 
@@ -662,8 +670,8 @@ function extractWineAttributes_2(a::T1, input::T2)::String where {T1<:agent, T2<
       keyword = Symbol(i * "_keyword")  # e.g. sweetness_keyword
       value = responsedict[keyword]
       if value != "NA" && !occursin(value, input)
-        errornote = "WARNING. Keyword $keyword: $value does not appear in the input. You must use information from the input only"
+        errornote = "In your previous attempt, keyword $keyword: $value does not appear in the input. You must use information from the input only"
-        println("Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
+        println("\nERROR YiemAgent extractWineAttributes_2() Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
         continue
       end
 
@@ -679,7 +687,7 @@ function extractWineAttributes_2(a::T1, input::T2)::String where {T1<:agent, T2<
       if !occursin("keyword", string(k))
         if v !== "NA" && (!occursin('-', v) || length(v) > 5)
           errornote = "WARNING: The non-range value {$k: $v} is not allowed. It should be specified in a range format, i.e. min-max."
-          println("Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
+          println("\nERROR YiemAgent extractWineAttributes_2() Attempt $attempt $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
           continue
         end
       end
@@ -755,7 +763,7 @@ function paraphrase(text2textInstructLLM::Function, text::String)
   for attempt in 1:10
     usermsg = """
               Text: $text
-              $errornote
+              P.S. $errornote
               """
 
     _prompt =
@@ -765,10 +773,11 @@ function paraphrase(text2textInstructLLM::Function, text::String)
       ]
 
     # put in model format
-    prompt = GeneralUtils.formatLLMtext(_prompt; formatname="qwen")
+    prompt = GeneralUtils.formatLLMtext(_prompt, "granite3")
 
     try
       response = text2textInstructLLM(prompt)
+      response = GeneralUtils.deFormatLLMtext(response, "granite3")
       # sometime the model response like this "here's how I would respond: ..."
       if occursin("respond:", response)
         errornote = "You don't need to intro your response"
@@ -780,13 +789,13 @@ function paraphrase(text2textInstructLLM::Function, text::String)
       response = replace(response, '`' => "")
       response = GeneralUtils.remove_french_accents(response)
 
-      # check whether response has all header
+      # check whether response has all answer's key points
       detected_kw = GeneralUtils.detect_keyword(header, response)
       if 0 ∈ values(detected_kw)
-        errornote = "\nYiemAgent paraphrase() response does not have all header"
+        errornote = "\nYiemAgent paraphrase() response does not have all answer's key points"
         continue
       elseif sum(values(detected_kw)) > length(header)
-        errornote = "\nnYiemAgent paraphrase() response has duplicated header"
+        errornote = "\nnYiemAgent paraphrase() response has duplicated answer's key points"
         continue
       end
 
@@ -984,7 +993,7 @@ end
 #   ]
 
 #   # put in model format
-#   prompt = GeneralUtils.formatLLMtext(_prompt; formatname="qwen")
+#   prompt = GeneralUtils.formatLLMtext(_prompt, "granite3")
 #   prompt *=
 #     """
 #     <|start_header_id|>assistant<|end_header_id|>

src/type.jl

@@ -9,11 +9,48 @@ using GeneralUtils
 
 abstract type agent end
 
-
 mutable struct companion <: agent
+  name::String  # agent name
   id::String    # agent id
-  systemmsg::Union{String, Nothing}
+  systemmsg::String # system message
+  tools::Dict  # tools
   maxHistoryMsg::Integer  # e.g. 21th and earlier messages will get summarized
+  chathistory::Vector{Dict{Symbol, Any}}
+  memory::Dict{Symbol, Any}
+  func::NamedTuple  # NamedTuple of functions 
+  llmFormatName::String
+end
+
+function companion(
+  func::NamedTuple  # NamedTuple of functions
+  ;
+  systemmsg::Union{String, Nothing}= nothing,
+  name::String= "Assistant",
+  id::String= GeneralUtils.uuid4snakecase(), 
+  maxHistoryMsg::Integer= 20,
+  chathistory::Vector{Dict{Symbol, String}} = Vector{Dict{Symbol, String}}(),
+  llmFormatName::String= "granite3"
+  )
+
+  if systemmsg === nothing
+    systemmsg = 
+      """
+      Your name: $name
+      Your sex: Female
+      Your role: You are a helpful assistant.
+      You should follow the following guidelines:
+      - Focus on the latest conversation.
+      - Your like to be short and concise.
+
+      Let's begin!
+      """
+  end
+
+  tools = Dict( # update input format
+      "CHATBOX"=> Dict(
+        :description => "- CHATBOX which you can use to talk with the user. The input is your intentions for the dialogue. Be specific.",
+      ),
+    )
 
   """ Memory
     Ref: Chat prompt format https://huggingface.co/TheBloke/Llama-2-7B-Chat-GGML/discussions/3
@@ -22,45 +59,31 @@ mutable struct companion <: agent
       Dict(:name=>"user", :text=> "Wassup!", :timestamp=> Dates.now()),
       Dict(:name=>"assistant", :text=> "Hi I'm your assistant.", :timestamp=> Dates.now()),
     ]
-
   """
-  chathistory::Vector{Dict{Symbol, Any}}
-  memory::Dict{Symbol, Any}
-
-  # communication function
-  text2textInstructLLM::Function
-end
-
-function companion(
-  text2textInstructLLM::Function
-  ;
-  id::String= string(uuid4()),
-  systemmsg::Union{String, Nothing}= nothing,
-  maxHistoryMsg::Integer= 20,
-  chathistory::Vector{Dict{Symbol, String}} = Vector{Dict{Symbol, String}}(),
-  )
-
   memory = Dict{Symbol, Any}(
-              :chatbox=> "",
+            :events=> Vector{Dict{Symbol, Any}}(),
-              :shortmem=> OrderedDict{Symbol, Any}(),
+            :state=> Dict{Symbol, Any}(), # state of the agent
-              :events=> Vector{Dict{Symbol, Any}}(),
+            :recap=> OrderedDict{Symbol, Any}(),  # recap summary of the conversation
-              :state=> Dict{Symbol, Any}(),
+          )
-            )
 
   newAgent = companion(
-                        id,
+    name,
-                        systemmsg,
+    id,
-                        maxHistoryMsg,
+    systemmsg,
-                        chathistory,
+    tools,
-                        memory,
+    maxHistoryMsg,
-                        text2textInstructLLM
+    chathistory,
-                      )
+    memory,
+    func,
+    llmFormatName
+    )
   
   return newAgent
 end
 
 
 
+
 """ A sommelier agent.
 
 # Arguments
@@ -134,19 +157,10 @@ mutable struct sommelier <: agent
   retailername::String
   tools::Dict   
   maxHistoryMsg::Integer  # e.g. 21th and earlier messages will get summarized
-
-  """ Memory
-    Ref: Chat prompt format https://huggingface.co/TheBloke/Llama-2-7B-Chat-GGML/discussions/3
-    NO "system" message in chathistory because I want to add it at the inference time
-    chathistory= [
-      Dict(:name=>"user", :text=> "Wassup!", :timestamp=> Dates.now()),
-      Dict(:name=>"assistant", :text=> "Hi I'm your assistant.", :timestamp=> Dates.now()),
-    ]
-
-  """
   chathistory::Vector{Dict{Symbol, Any}}
   memory::Dict{Symbol, Any}
   func  # NamedTuple of functions 
+  llmFormatName::String
 end
 
 function sommelier(
@@ -157,6 +171,7 @@ function sommelier(
   retailername::String= "retailer_name",
   maxHistoryMsg::Integer= 20,
   chathistory::Vector{Dict{Symbol, String}} = Vector{Dict{Symbol, String}}(),
+  llmFormatName::String= "granite3"
   )
 
   tools = Dict( # update input format
@@ -170,16 +185,17 @@ function sommelier(
         :input => """<input>Input is a JSON-formatted string that contains a detailed and precise search query.</input><input example>{\"wine type\": \"rose\", \"price\": \"max 35\", \"sweetness level\": \"sweet\", \"intensity level\": \"light bodied\", \"Tannin level\": \"low\", \"Acidity level\": \"low\"}</input example>""",
         :output => """<output>Output are wines that match the search query in JSON format.""",
       ),
-      # "finalanswer"=> Dict(
-      #   :description => "<tool description>Useful for when you are ready to recommend wines to the user.</tool description>",
-      #   :input => """<input format>{\"finalanswer\": \"some text\"}.</input format><input example>{\"finalanswer\": \"I recommend Zena Crown Vista\"}</input example>""",
-      #   :output => "" ,
-      #   :func => nothing,
-      # ),
     )
 
+  """ Memory
+    Ref: Chat prompt format https://huggingface.co/TheBloke/Llama-2-7B-Chat-GGML/discussions/3
+    NO "system" message in chathistory because I want to add it at the inference time
+    chathistory= [
+      Dict(:name=>"user", :text=> "Wassup!", :timestamp=> Dates.now()),
+      Dict(:name=>"assistant", :text=> "Hi I'm your assistant.", :timestamp=> Dates.now()),
+    ]
+  """
   memory = Dict{Symbol, Any}(
-            :chatbox=> "",
             :shortmem=> OrderedDict{Symbol, Any}(
               :available_wine=> [],
               :found_wine=> [], # used by decisionMaker(). This is to prevent decisionMaker() keep presenting the same wines
@@ -198,7 +214,8 @@ function sommelier(
     maxHistoryMsg,
     chathistory,
     memory,
-    func
+    func,
+    llmFormatName
     )
   
   return newAgent

src/util.jl

@@ -154,11 +154,11 @@ function chatHistoryToText(vecd::Vector; withkey=true, range=nothing)::String
     # Loop through each dictionary in the input vector
     for d in elements
         # Extract the 'name' and 'text' keys from the dictionary
-        name = d[:name]
+        name = titlecase(d[:name])
         _text = d[:text]
         
         # Append the formatted string to the text variable
-        text *= "$name:> $_text \n"
+        text *= "$name> $_text \n"
       end
   else
     # Loop through each dictionary in the input vector
@@ -239,19 +239,22 @@ function eventdict(;
   outcome::Union{String, Nothing}=nothing,
   note::Union{String, Nothing}=nothing,
   )
-  return Dict{Symbol, Any}(
+
-      :event_description=> event_description,
+  d = Dict{Symbol, Any}(
-      :timestamp=> timestamp,
+    :event_description=> event_description,
-      :subject=> subject,
+    :timestamp=> timestamp,
-      :thought=> thought,
+    :subject=> subject,
-      :actionname=> actionname,
+    :thought=> thought,
-      :actioninput=> actioninput,
+    :actionname=> actionname,
-      :location=> location,
+    :actioninput=> actioninput,
-      :equipment_used=> equipment_used,
+    :location=> location,
-      :material_used=> material_used,
+    :equipment_used=> equipment_used,
-      :outcome=> outcome,
+    :material_used=> material_used,
-      :note=> note,
+    :outcome=> outcome,
-    )
+    :note=> note,
+  )
+
+  return d
 end
 
 
@@ -310,221 +313,34 @@ function createTimeline(events::T1; eventindex::Union{UnitRange, Nothing}=nothin
 end
 
 
+# function createTimeline(events::T1; eventindex::Union{UnitRange, Nothing}=nothing
+#   ) where {T1<:AbstractVector}
+#   # Initialize empty timeline string
+#   timeline = ""
+  
+#   # Determine which indices to use - either provided range or full length
+#   ind = 
+#     if eventindex !== nothing
+#       [eventindex...]
+#     else
+#       1:length(events)
+#     end
 
-# """ Convert a single chat dictionary into LLM model instruct format.
+#   # Iterate through events and format each one
-
+#   for (i, event) in zip(ind, events)
-# # Llama 3 instruct format example
+#     # If no outcome exists, format without outcome
-#   <|system|>
+#     subject = titlecase(event[:subject])
-#   You are a helpful AI assistant.<|end|>
+#     if event[:outcome] === nothing
-#   <|user|>
+      
-#   I am going to Paris, what should I see?<|end|>
+#       timeline *= "Event_$i) Who: $subject Action_name: $(event[:actionname]) Action_input: $(event[:actioninput])\n"
-#   <|assistant|>
+#     # If outcome exists, include it in formatting  
-#   Paris, the capital of France, is known for its stunning architecture, art museums."<|end|>
+#     else
-#   <|user|>
+#       timeline *= "Event_$i) Who: $subject Action_name: $(event[:actionname]) Action_input: $(event[:actioninput]) Action output: $(event[:outcome])\n"
-#   What is so great about #1?<|end|>
-#   <|assistant|>
-
-
-# # Arguments
-#   - `name::T`
-#     message owner name e.f. "system", "user" or "assistant"
-#   - `text::T`
-
-# # Return
-#   - `formattedtext::String`
-#     text formatted to model format
-
-# # Example
-# ```jldoctest
-# julia> using Revise
-# julia> using YiemAgent
-# julia> d = Dict(:name=> "system",:text=> "You are a helpful, respectful and honest assistant.",)
-# julia> formattedtext = YiemAgent.formatLLMtext_phi3instruct(d[:name], d[:text])
-
-# ```
-
-# Signature
-# """
-# function formatLLMtext_phi3instruct(name::T, text::T) where {T<:AbstractString}
-#   formattedtext = 
-#   """
-#   <|$name|>
-#   $text<|end|>\n
-#   """
-
-#   return formattedtext
-# end
-
-
-# """ Convert a single chat dictionary into LLM model instruct format.
-
-# # Llama 3 instruct format example
-#   <|begin_of_text|>
-#   <|start_header_id|>system<|end_header_id|>
-#   You are a helpful assistant.
-#   <|eot_id|>
-#   <|start_header_id|>user<|end_header_id|>
-#   Get me an icecream.
-#   <|eot_id|>
-#   <|start_header_id|>assistant<|end_header_id|>
-#   Go buy it yourself at 7-11.
-#   <|eot_id|>
-
-# # Arguments
-#   - `name::T`
-#     message owner name e.f. "system", "user" or "assistant"
-#   - `text::T`
-
-# # Return
-#   - `formattedtext::String`
-#     text formatted to model format
-
-# # Example
-# ```jldoctest
-# julia> using Revise
-# julia> using YiemAgent
-# julia> d = Dict(:name=> "system",:text=> "You are a helpful, respectful and honest assistant.",)
-# julia> formattedtext = YiemAgent.formatLLMtext_llama3instruct(d[:name], d[:text])
-# "<|begin_of_text|>\n  <|start_header_id|>system<|end_header_id|>\n  You are a helpful, respectful and honest assistant.\n  <|eot_id|>\n"
-# ```
-
-# Signature
-# """
-# function formatLLMtext_llama3instruct(name::T, text::T) where {T<:AbstractString}
-#   formattedtext = 
-#   if name == "system"
-#     """
-#     <|begin_of_text|>
-#     <|start_header_id|>$name<|end_header_id|>
-#     $text
-#     <|eot_id|>
-#     """
-#   else
-#     """
-#     <|start_header_id|>$name<|end_header_id|>
-#     $text
-#     <|eot_id|>
-#     """
-#   end
-
-#   return formattedtext
-# end
-
-
-
-# """ Convert a chat messages in vector of dictionary into LLM model instruct format.
-
-# # Arguments
-#   - `messages::Vector{Dict{Symbol, T}}`
-#     message owner name e.f. "system", "user" or "assistant"
-#   - `formatname::T`
-#     format name to be used
-
-# # Return
-#   - `formattedtext::String`
-#     text formatted to model format
-
-# # Example
-# ```jldoctest
-# julia> using Revise
-# julia> using YiemAgent
-# julia> chatmessage = [
-#           Dict(:name=> "system",:text=> "You are a helpful, respectful and honest assistant.",),
-#           Dict(:name=> "user",:text=> "list me all planets in our solar system.",),
-#           Dict(:name=> "assistant",:text=> "I'm sorry. I don't know. You tell me.",),
-#         ]
-# julia> formattedtext = YiemAgent.formatLLMtext(chatmessage, "llama3instruct")
-# "<|begin_of_text|>\n  <|start_header_id|>system<|end_header_id|>\n  You are a helpful, respectful and honest assistant.\n  <|eot_id|>\n  <|start_header_id|>user<|end_header_id|>\n  list me all planets in our solar system.\n  <|eot_id|>\n  <|start_header_id|>assistant<|end_header_id|>\n  I'm sorry. I don't know. You tell me.\n  <|eot_id|>\n"
-# ```
-
-# # Signature
-# """
-# function formatLLMtext(messages::Vector{Dict{Symbol, T}}, 
-#                       formatname::String="llama3instruct") where {T<:Any}
-#   f = if formatname == "llama3instruct"
-#     formatLLMtext_llama3instruct
-#   elseif formatname == "mistral"
-#     # not define yet
-#   elseif formatname == "phi3instruct"
-#     formatLLMtext_phi3instruct
-#   else
-#     error("$formatname template not define yet")
-#   end
-
-#   str = ""
-#   for t in messages
-#     str *= f(t[:name], t[:text])
-#   end
-
-#   # add <|assistant|> so that the model don't generate it and I don't need to clean it up later
-#   if formatname == "phi3instruct"
-#     str *= "<|assistant|>\n"
-#   end
-
-#   return str
-# end
-
-
-# """
-
-#   Arguments\n
-#   -----
-    
-#   Return\n
-#   -----
-
-#   Example\n
-#   -----
-#   ```jldoctest
-#   julia> 
-#   ```
-
-#   TODO\n
-#   -----
-#     [] update docstring
-#     [PENDING] implement the function
-
-#   Signature\n
-#   -----
-# """
-# function iterativeprompting(a::T, prompt::String, verification::Function) where {T<:agent}
-#   msgMeta = GeneralUtils.generate_msgMeta(
-#     a.config[:externalService][:text2textinstruct],
-#     senderName= "iterativeprompting",
-#     senderId= a.id,
-#     receiverName= "text2textinstruct",
-#   )
-
-#   outgoingMsg = Dict(
-#     :msgMeta=> msgMeta,
-#     :payload=> Dict(
-#       :text=> prompt,
-#     )
-#   )
-
-#   success = nothing
-#   result = nothing
-#   critique = ""
-
-#   # iteration loop
-#   while true
-#     # send prompt to LLM
-#     response = GeneralUtils.sendReceiveMqttMsg(outgoingMsg)
-#     error("--> iterativeprompting")
-#     # check for correctness and get feedback
-#     success, _critique = verification(response)
-
-#     if success
-#       result = response
-#       break
-#     else  
-#       # add critique to prompt
-#       critique *= _critique * "\n"
-#       replace!(prompt, "Critique: ..." => "Critique: $critique")
 #     end
 #   end
-  
+
-#   return (success=success, result=result)
+#   # Return formatted timeline string
+#   return timeline
 # end
 
 
@@ -555,11 +371,6 @@ end
 
 
 
-
-
-
-
-
 
 
 

test/test_1.jl

@@ -36,7 +36,12 @@ function executeSQLVectorDB(sql)
   return result
 end
 
-function text2textInstructLLM(prompt::String; maxattempt::Integer=2, modelsize::String="medium")
+function text2textInstructLLM(prompt::String; maxattempt::Integer=3, modelsize::String="medium",
+    llmkwargs=Dict(
+      :num_ctx => 32768,
+      :temperature => 0.1,
+    )
+  )
   msgMeta = GeneralUtils.generate_msgMeta(
     config[:externalservice][:loadbalancer][:mqtttopic];
     msgPurpose="inference",
@@ -51,10 +56,7 @@ function text2textInstructLLM(prompt::String; maxattempt::Integer=2, modelsize::
     :msgMeta => msgMeta,
     :payload => Dict(
       :text => prompt,
-      :kwargs => Dict(
+      :kwargs => llmkwargs
-        :num_ctx => 16384,
-        :temperature => 0.2,
-      )
     )
   )
 
@@ -195,7 +197,7 @@ function insertSommelierDecision(recentevents::T1, decision::T2; maxdistance::In
   row, col = size(df)
   distance = row == 0 ? Inf : df[1, :distance]
   if row == 0 || distance > maxdistance # no close enough SQL stored in the database
-    recentevents_embedding = a.func[:getEmbedding](recentevents)[1]
+    recentevents_embedding = getEmbedding(recentevents)[1]
     recentevents = replace(recentevents, "'" => "")
     decision_json = JSON3.write(decision)
     decision_base64 = base64encode(decision_json)