Compare commits
28 Commits
c085a445e2
...
main
| Author | SHA1 | Date | |
|---|---|---|---|
| 66c86b177c | |||
| cf71f56771 | |||
| 0fcf8c8669 | |||
| 388b4716e6 | |||
| c27a7008fe | |||
| 480ecf3e74 | |||
| c80c4ca65d | |||
| 57a8e9cafc | |||
| 831014cabf | |||
| 243e3fe10b | |||
| c594a34e4e | |||
| 2efb016646 | |||
| 216a8bdabb | |||
| cbaa480e84 | |||
| 6cb4073e29 | |||
| 42b8f5bdb1 | |||
| 35f1482228 | |||
| b4cac4f383 | |||
| 685ee7a48f | |||
| b55ae31e5b | |||
| e8e1764bb4 | |||
| e0b3ffa8e3 | |||
| ec9f44e5a1 | |||
| 6b3f8620e1 | |||
| 9cd37317d7 | |||
| 1577d7ae25 | |||
| c4e255ec2a | |||
| 681a91a0ca |
+22
-15
@@ -2,7 +2,7 @@
|
||||
|
||||
julia_version = "1.12.6"
|
||||
manifest_format = "2.0"
|
||||
project_hash = "6e0efa362e5748de9ce219e0376be2f93d464376"
|
||||
project_hash = "f82baf5953223c6185bd47af518fd540515402d1"
|
||||
|
||||
[[deps.Accessors]]
|
||||
deps = ["CompositionsBase", "ConstructionBase", "Dates", "InverseFunctions", "MacroTools"]
|
||||
@@ -268,19 +268,24 @@ deps = ["Random"]
|
||||
uuid = "9fa8497b-333b-5362-9e8d-4d0656e87820"
|
||||
version = "1.11.0"
|
||||
|
||||
[[deps.Gamma]]
|
||||
git-tree-sha1 = "86f86b6168a016ed88e4ae4e64577b98c3b59e8e"
|
||||
uuid = "a0844989-3bd2-4988-8bea-c9407ab0941b"
|
||||
version = "1.1.0"
|
||||
|
||||
[[deps.GeneralUtils]]
|
||||
deps = ["CSV", "DataFrames", "DataStructures", "Dates", "Distributions", "JSON", "NATS", "PrettyPrinting", "Random", "Revise", "SHA", "UUIDs"]
|
||||
git-tree-sha1 = "f1bad7621e6ac2d235adc8c593b0aff87bd1d93e"
|
||||
deps = ["CSV", "DataFrames", "DataStructures", "Dates", "Distributions", "HTTP", "JSON", "NATS", "PrettyPrinting", "Random", "Revise", "SHA", "UUIDs"]
|
||||
git-tree-sha1 = "7c0600c166a5deb2c607018a491c04eb25969c2e"
|
||||
repo-rev = "main"
|
||||
repo-url = "https://git.yiem.cc/ton/GeneralUtils"
|
||||
uuid = "c6c72f09-b708-4ac8-ac7c-2084d70108fe"
|
||||
version = "0.4.2"
|
||||
version = "0.4.9"
|
||||
|
||||
[[deps.HTTP]]
|
||||
deps = ["Base64", "CodecZlib", "Dates", "EnumX", "PrecompileTools", "Random", "Reseau", "SHA", "URIs", "UUIDs", "Zlib_jll"]
|
||||
git-tree-sha1 = "69343dd8afb1671b84c3aa2dda511238d0919a55"
|
||||
git-tree-sha1 = "eda1d37cb55d90a17d0957c75841138c88b361a1"
|
||||
uuid = "cd3eb016-35fb-5094-929b-558a96fad6f3"
|
||||
version = "2.5.0"
|
||||
version = "2.5.4"
|
||||
|
||||
[[deps.HashArrayMappedTries]]
|
||||
git-tree-sha1 = "2eaa69a7cab70a52b9687c8bf950a5a93ec895ae"
|
||||
@@ -288,10 +293,10 @@ uuid = "076d061b-32b6-4027-95e0-9a2c6f6d7e74"
|
||||
version = "0.2.0"
|
||||
|
||||
[[deps.HypergeometricFunctions]]
|
||||
deps = ["LinearAlgebra", "OpenLibm_jll", "SpecialFunctions"]
|
||||
git-tree-sha1 = "68c173f4f449de5b438ee67ed0c9c748dc31a2ec"
|
||||
deps = ["Gamma", "LinearAlgebra"]
|
||||
git-tree-sha1 = "18d7deab5fb0440dc6a7b6993c5c27b25420de10"
|
||||
uuid = "34004b35-14d8-5ef3-9330-4cdb6864b03a"
|
||||
version = "0.3.28"
|
||||
version = "0.3.29"
|
||||
|
||||
[[deps.ICU_jll]]
|
||||
deps = ["Artifacts", "JLLWrappers", "Libdl"]
|
||||
@@ -401,9 +406,11 @@ version = "1.21.3+0"
|
||||
|
||||
[[deps.LLMMCTS]]
|
||||
deps = ["GeneralUtils", "JSON", "PrettyPrinting"]
|
||||
path = "../LLMMCTS"
|
||||
git-tree-sha1 = "6b4f123b03c0fcce5b21c0dbcb947e8dd23f333a"
|
||||
repo-rev = "main"
|
||||
repo-url = "https://git.yiem.cc/ton/LLMMCTS"
|
||||
uuid = "d76c5a4d-449e-4835-8cc4-dd86ec44f241"
|
||||
version = "0.1.4"
|
||||
version = "0.1.5"
|
||||
|
||||
[[deps.LaTeXStrings]]
|
||||
git-tree-sha1 = "dda21b8cbd6a6c40d9d02a73230f9d70fed6918c"
|
||||
@@ -727,9 +734,9 @@ version = "0.5.1+0"
|
||||
|
||||
[[deps.Roots]]
|
||||
deps = ["Accessors", "CommonSolve", "Printf"]
|
||||
git-tree-sha1 = "91cfb1cb4f6e27557cc2df798a31eff6089a41eb"
|
||||
git-tree-sha1 = "ed45bcc7cf3c8887595b973f2b1efbe91dcc50ec"
|
||||
uuid = "f2b01f46-fcfa-551c-844a-d8ac1e96c665"
|
||||
version = "3.0.0"
|
||||
version = "3.0.1"
|
||||
|
||||
[deps.Roots.extensions]
|
||||
RootsChainRulesCoreExt = "ChainRulesCore"
|
||||
@@ -752,10 +759,10 @@ uuid = "ea8e919c-243c-51af-8825-aaa63cd721ce"
|
||||
version = "0.7.0"
|
||||
|
||||
[[deps.SQLLLM]]
|
||||
deps = ["CSV", "DataFrames", "DataStructures", "Dates", "FileIO", "GeneralUtils", "HTTP", "JSON", "LLMMCTS", "LibPQ", "PrettyPrinting", "Random", "Revise", "StatsBase", "Tables", "URIs", "UUIDs"]
|
||||
deps = ["CSV", "DataFrames", "DataStructures", "Dates", "FileIO", "GeneralUtils", "HTTP", "JSON", "LibPQ", "PrettyPrinting", "Random", "Revise", "StatsBase", "Tables", "URIs", "UUIDs"]
|
||||
path = "."
|
||||
uuid = "2ebc79c7-cc10-4a3a-9665-d2e1d61e63d3"
|
||||
version = "0.2.4"
|
||||
version = "0.2.5"
|
||||
|
||||
[[deps.SQLStrings]]
|
||||
git-tree-sha1 = "55de0530689832b1d3d43491ee6b67bd54d3323c"
|
||||
|
||||
+3
-2
@@ -1,6 +1,6 @@
|
||||
name = "SQLLLM"
|
||||
uuid = "2ebc79c7-cc10-4a3a-9665-d2e1d61e63d3"
|
||||
version = "0.2.4"
|
||||
version = "0.2.8"
|
||||
authors = ["narawat lamaiin <narawat@outlook.com>"]
|
||||
|
||||
[deps]
|
||||
@@ -24,5 +24,6 @@ UUIDs = "cf7118a7-6976-5b1a-9a39-7adc72f591a4"
|
||||
|
||||
[compat]
|
||||
Dates = "1.11.0"
|
||||
GeneralUtils = "0.4.2"
|
||||
GeneralUtils = "0.4.0 - 0.9.0"
|
||||
JSON = "1.6.1"
|
||||
LLMMCTS = "0.1.5 - 0.9.0"
|
||||
|
||||
@@ -0,0 +1,22 @@
|
||||
d = OrderedDict{String, Any}("1" => Dict{String, Any}("action_name" => "RUNSQL", "action_input" => "SELECT w.winery, w.wine_name, w.wine_id, w.vintage, w.region, w.country, w.wine_type, w.grape, w.serving_temperature, w.sweetness, w.intensity, w.tannin, w.acidity, w.tasting_notes, rw.price, rw.currency, w.image_url FROM wine w JOIN retailer_wine rw ON w.wine_id = rw.wine_id JOIN retailer r ON rw.retailer_id = r.retailer_id WHERE w.wine_name ILIKE 'Montrachet Grand Cru' AND w.winery ILIKE 'Domaine Jacques Prieur' AND w.region ILIKE 'Montrachet' AND r.retailer_id = 'f54eab6b-7650-4448-b009-c53f3efbcc3b';", "action_result" => "No records found. Try loosening your search criteria.", "plan" => "The user wants to retrieve specific details about a wine, including its characteristics and pricing at a specific retailer, based on several criteria: wine name, winery, region, retailer name, and retailer ID. I need to join the 'wine', 'retailer_wine', and 'retailer' tables. I will use the wine attributes (name, winery, region) and the retailer ID to filter the results and select all requested fields, using ILIKE for text matching as per guidelines."), "2" => Dict{String, Any}("action_name" => "RUNSQL", "action_input" => "SELECT retailer_name FROM retailer WHERE retailer_id = 'f54eab6b-7650-4448-b009-c53f3efbcc3b';", "action_result" => "1) retailer_name: Yiem Wines Ltd\n", "plan" => "The previous query failed with no records found, indicating that the combination of all constraints (wine name, winery, region, and retailer ID) is too restrictive. To proceed, I will first verify the retailer's name using the provided ID. Then, I will execute a new query, joining the 'wine', 'retailer_wine', and 'retailer' tables, filtering only by the wine name ('Montrachet Grand Cru') and the specific retailer ID ('f54eab6b-7650-4448-b009-c53f3efbcc3b'), removing the 'winery' and 'region' constraints to see if any matching records exist at that retailer."))
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
+341
-654
File diff suppressed because it is too large
Load Diff
+123
-124
@@ -481,20 +481,9 @@ julia> response = SQLLLM.SQLexecution(executeSQL, sql)
|
||||
# Signature
|
||||
"""
|
||||
function SQLexecution(executeSQL::Function, sql::T
|
||||
) where {T<:AbstractString}
|
||||
)::NamedTuple where {T<:AbstractString}
|
||||
|
||||
try
|
||||
#XXX dummy SQL. use for testing
|
||||
# sql = "SELECT w.wine_name FROM wine w JOIN wine_food wf ON w.wine_id = wf.wine_id JOIN food f ON wf.food_id = f.food_id WHERE f.\"food_name\" = 'lamb';"
|
||||
# sql = " SELECT w.wine_name FROM wine w JOIN food f ON f.food_name = 'lamb' JOIN wine_food wf ON w.wine_id = wf.wine_id AND f.food_id = wf.food_id GROUP BY w.wine_name ORDER BY COUNT(DISTINCT w.wine_id) DESC;"
|
||||
# sql = " SELECT COUNT(DISTINCT wf.wine_id) FROM wine w JOIN wine_food wf ON w.wine_id = wf.wine_id JOIN food f ON wf.food_id = f.food_id WHERE f.food_name ILIKE '%lamb%'"
|
||||
|
||||
#XXX use for package testing, remove when done
|
||||
# ans = "1.schilfwein zweigelt 2.cabernet sauvignon reserve limited edition"
|
||||
# ans = "There are 1500 wines that can be paired with lamb."
|
||||
# ans = "1500"
|
||||
# return (response=ans, errormsg=nothing, reward=1, isterminal=true)
|
||||
|
||||
# add LIMIT to the SQL to prevent loading large data
|
||||
sql = strip(sql)
|
||||
|
||||
@@ -508,39 +497,36 @@ function SQLexecution(executeSQL::Function, sql::T
|
||||
else
|
||||
sql = sql * ";"
|
||||
end
|
||||
println("\n~~~ SQLexecution() SQL: ", @__FILE__, " ", @__LINE__)
|
||||
println(sql)
|
||||
|
||||
result = executeSQL(sql)
|
||||
df = DataFrame(result)
|
||||
|
||||
tablesize = size(df)
|
||||
row, column = tablesize
|
||||
if row == 0
|
||||
error("\nThe resulting table has 0 row. Please try again.")
|
||||
return (result_str="No records found. Try loosening your search criteria.", result_raw=nothing, success=true, errormsg=nothing)
|
||||
elseif column > 30
|
||||
error("\nSQL execution failed. An unexpected error occurred. Please try again.")
|
||||
return (result_str="There are more than 30 columns. Please be more specific.", result_raw=df, success=true, errormsg=nothing)
|
||||
else
|
||||
df1 =
|
||||
if row > 2
|
||||
# ramdom row to pick
|
||||
df[sample(1:nrow(df), 2, replace=false), :] # random select 2 rows from df
|
||||
else
|
||||
df
|
||||
end
|
||||
result = GeneralUtils.dfToString(df1)
|
||||
# println("\n~~~ SQLexecution() result: ", @__FILE__, " ", @__LINE__)
|
||||
# println(sql)
|
||||
# println(df1)
|
||||
# println("\n")
|
||||
return (result_str=result, result_raw=df1, success=true, errormsg=nothing)
|
||||
end
|
||||
|
||||
df1 =
|
||||
if row > 2
|
||||
# ramdom row to pick
|
||||
df[sample(1:nrow(df), 2, replace=false), :] # random select 2 rows from df
|
||||
else
|
||||
df
|
||||
end
|
||||
|
||||
println("\n~~~ SQLexecution() result: ", @__FILE__, " ", @__LINE__)
|
||||
println(df1)
|
||||
return (result=df1, success=true, errormsg=nothing)
|
||||
catch e
|
||||
io = IOBuffer()
|
||||
showerror(io, e)
|
||||
errorMsg = String(take!(io))
|
||||
st = sprint((io, v) -> show(io, "text/plain", v), stacktrace(catch_backtrace()))
|
||||
println(errorMsg)
|
||||
response = (result=nothing, success=false, errormsg=errorMsg)
|
||||
return response
|
||||
return (result_str=nothing, result_raw=nothing, success=false, errormsg=errorMsg)
|
||||
end
|
||||
end
|
||||
|
||||
@@ -559,7 +545,7 @@ end
|
||||
- `result::String`
|
||||
|
||||
# Signature
|
||||
"""
|
||||
""" #PENDING
|
||||
function extractContent_dataframe(df::DataFrame, text2textInstructLLM::Function, action::String,
|
||||
llmFormatName::String
|
||||
)::String
|
||||
@@ -633,7 +619,7 @@ function extractContent_dataframe(df::DataFrame, text2textInstructLLM::Function,
|
||||
dictkey = ["about_resulting_table", "search_summary"]
|
||||
|
||||
for i in 1:5
|
||||
response = text2textInstructLLM(prompt, modelsize="medium")
|
||||
response = text2textInstructLLM("ramdom_id", prompt)
|
||||
response = GeneralUtils.deFormatLLMtext(response, llmFormatName)
|
||||
think, response = GeneralUtils.extractthink(response)
|
||||
|
||||
@@ -653,7 +639,6 @@ function extractContent_dataframe(df::DataFrame, text2textInstructLLM::Function,
|
||||
responsedict = GeneralUtils.textToDict(response, header;
|
||||
dictKey=dictkey, symbolkey=false)
|
||||
|
||||
# result = dfstr
|
||||
result =
|
||||
"""
|
||||
Summary: $(responsedict["search_summary"])
|
||||
@@ -821,142 +806,156 @@ julia>
|
||||
|
||||
# Notes
|
||||
- The function makes up to 10 attempts to get a valid response from the LLM
|
||||
- Each state in highValueStateList should contain a thoughtHistory with action_input and observation
|
||||
- Each state in highValueStateList should contain a action_history with action_input and observation
|
||||
- The LLM evaluates attempts based on accuracy and relevance to the original question
|
||||
"""
|
||||
function compareState(question::String, highValueStateList::Vector{T},
|
||||
text2textInstructLLM::Function, llmFormatName::String
|
||||
text2textInstructLLM::Function; maxattempt::Integer=10
|
||||
)::Integer where {T<:AbstractDict}
|
||||
|
||||
systemmsg =
|
||||
"""
|
||||
Your profile:
|
||||
- You are a helpful assistant
|
||||
Situation:
|
||||
- The user has made multiple attempts to solve the question, resulting in various answers
|
||||
Your mission:
|
||||
- Identify and select the most accurate and relevant response from these multiple results for the user
|
||||
At each round of conversation, you will be given the following:
|
||||
Question: the question the user is trying to answer
|
||||
Attempt: the user's attempted actions and their corresponding results
|
||||
You should then respond to the user with the following:
|
||||
Comparison: detailed comparison of all results from all attempts from various aspects.
|
||||
Rationale: a brief explanation of why the selected response is the most accurate and relevant
|
||||
Selected_response_number: the number the selected response in the list of results (e.g., 1, 2, 3, ...)
|
||||
You should only respond in format as described below:
|
||||
Comparison: ...
|
||||
Rationale: ...
|
||||
Selected_response_number: ...
|
||||
Here are some examples:
|
||||
User's question: "How many German wines do you have?"
|
||||
Attempt 1)
|
||||
Action: SELECT COUNT(*) FROM wines WHERE country = 'Germany'
|
||||
Result: 100 wines
|
||||
Attempt 2)
|
||||
Action: SELECT COUNT(*) FROM wines WHERE country = 'Germany' AND type = 'Red'
|
||||
Result: 50 red wines
|
||||
Comparison: The second attempt counts only German red wines while the first attempt includes all German wines.
|
||||
Rationale: The user is asking for the number of German wines without specifying a type, so the most accurate response is the first attempt because it includes all German wines.
|
||||
Selected_response_number:1
|
||||
# Your profile:
|
||||
- You are a helpful assistant
|
||||
|
||||
Let's begin!
|
||||
# Situation:
|
||||
- The user has made multiple attempts to solve the question, resulting in various answers
|
||||
|
||||
# Your mission:
|
||||
- Identify and select the most accurate and relevant response from these multiple results for the user
|
||||
|
||||
# At each round of conversation, you will be given the following:
|
||||
Question: the question the user is trying to answer
|
||||
Attempt: the user's attempted actions and their corresponding results
|
||||
|
||||
# You should then respond to the user with the following:
|
||||
1) "comparison", detailed comparison of all results from all attempts from various aspects.
|
||||
2) "rationale", a brief explanation of why the selected response is the most accurate and relevant
|
||||
3) "selected_response_number", the number the selected response in the list of results (e.g., 1, 2, 3, ...)
|
||||
|
||||
# you should only respond in JSON format as described below
|
||||
"comparison": "..."
|
||||
"rationale": "..."
|
||||
"selected_response_number": "..."
|
||||
|
||||
# Here are some examples:
|
||||
Question: "How many German wines do you have?"
|
||||
Attempt 1)
|
||||
action_name: RUNSQL
|
||||
action_input: SELECT COUNT(*) FROM wines WHERE country = 'Germany'
|
||||
action_result: 100 wines
|
||||
Attempt 2)
|
||||
action_name: RUNSQL
|
||||
action_input: SELECT COUNT(*) FROM wines WHERE country = 'Germany' AND type = 'Red'
|
||||
action_result: 50 red wines
|
||||
|
||||
"comparison": "The second attempt counts only German red wines while the first attempt includes all German wines."
|
||||
"rationale": "The user is asking for the number of German wines without specifying a type, so the most accurate response is the first attempt because it includes all German wines."
|
||||
"selected_response_number": "1"
|
||||
"""
|
||||
|
||||
potentialSolution = []
|
||||
keys = ["action_input", "observation"]
|
||||
requiredKeys = ["comparison", "rationale", "selected_response_number"]
|
||||
|
||||
potentialSolution = []
|
||||
includekeys = ["action_name", "action_input", "action_result"]
|
||||
# extract the last action_name, action_input, observation of each state in highValueStateList and store them in a dictionary then push into potentialSolution
|
||||
for state in highValueStateList
|
||||
thoughtHistory = state["thoughtHistory"]
|
||||
_, currentstate_latestIndice =
|
||||
GeneralUtils.findHighestIndexKey(thoughtHistory, keys[1])
|
||||
latestKeys = makekey.(keys, currentstate_latestIndice)
|
||||
action_history = state["action_history"]
|
||||
latestKeys = [i for i in keys(action_history)][end]
|
||||
d = Dict()
|
||||
# get the last action_name, action_input, observation of currentstate
|
||||
for (i,v) in enumerate(keys)
|
||||
d[v] = thoughtHistory[latestKeys[i]]
|
||||
for (i,v) in enumerate(includekeys)
|
||||
latest_action = action_history[latestKeys]
|
||||
d[v] = latest_action[v]
|
||||
end
|
||||
push!(potentialSolution, d)
|
||||
end
|
||||
|
||||
println("\n")
|
||||
@show potentialSolution
|
||||
println("--- ", @__FILE__, @__LINE__)
|
||||
|
||||
"""
|
||||
# put potential solutions from potentialSolution into the following form
|
||||
Attempt 1)
|
||||
action_name:
|
||||
action_input:
|
||||
observation:
|
||||
action_result:
|
||||
Attempt 2)
|
||||
action_name:`
|
||||
action_name:
|
||||
action_input:
|
||||
observation:`
|
||||
action_result:
|
||||
...
|
||||
"""
|
||||
|
||||
potentialSolutionStr = ""
|
||||
for (i, state) in enumerate(potentialSolution)
|
||||
potentialSolutionStr *= "Attempt $i)\n"
|
||||
for k in keys
|
||||
for k in includekeys
|
||||
potentialSolutionStr *= "$k: $(state[k])\n"
|
||||
println("")
|
||||
end
|
||||
end
|
||||
|
||||
errornote = "N/A"
|
||||
usermsg =
|
||||
"""
|
||||
Question: $question
|
||||
$potentialSolutionStr
|
||||
"""
|
||||
|
||||
for attempt in 1:10
|
||||
errorFlag = false
|
||||
msg = Dict(
|
||||
"model" => "gemma-4-E4B-it-UD-Q4_K_XL",
|
||||
"messages" => [
|
||||
Dict(
|
||||
"role" => "system",
|
||||
"content" => [
|
||||
Dict("type" => "text", "text" => systemmsg),
|
||||
]
|
||||
),
|
||||
Dict(
|
||||
"role" => "user",
|
||||
"content" => [
|
||||
Dict("type" => "text", "text" => usermsg),
|
||||
]
|
||||
),
|
||||
],
|
||||
"temperature" => 0.7
|
||||
)
|
||||
|
||||
usermsg =
|
||||
"""
|
||||
Question: $question
|
||||
Attempts: $potentialSolutionStr
|
||||
P.S. $errornote
|
||||
"""
|
||||
|
||||
_prompt =
|
||||
[
|
||||
Dict(:name=> "system", :text=> systemmsg),
|
||||
Dict(:name=> "user", :text=> usermsg)
|
||||
]
|
||||
|
||||
# put in model format
|
||||
prompt = GeneralUtils.formatLLMtext(_prompt, llmFormatName)
|
||||
|
||||
header = ["Comparison:", "Rationale:", "Selected_response_number:"]
|
||||
dictkey = ["comparison", "rationale", "selected_response_number"]
|
||||
|
||||
response = text2textInstructLLM(prompt, modelsize="medium")
|
||||
|
||||
# sometime LLM output something like **Comprehension**: which is not expected
|
||||
response = replace(response, "**"=>"")
|
||||
response = replace(response, "***"=>"")
|
||||
response = GeneralUtils.deFormatLLMtext(response, llmFormatName)
|
||||
for attempt in 1:maxattempt
|
||||
response = text2textInstructLLM("random_id", msg)
|
||||
response = GeneralUtils.clean_json_response(response)
|
||||
response = GeneralUtils.remove_french_accents(response)
|
||||
think, response = GeneralUtils.extractthink(response)
|
||||
|
||||
# check whether response has all header
|
||||
detected_kw = GeneralUtils.detectKeywordVariation(header, response)
|
||||
missingkeys = [k for (k, v) in detected_kw if v === nothing]
|
||||
if !isempty(missingkeys)
|
||||
errornote = "$missingkeys are missing from your previous response"
|
||||
println("\nERROR SQLLLM extractContent_dataframe() $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
|
||||
continue
|
||||
elseif sum([length(i) for i in values(detected_kw)]) > length(header)
|
||||
errornote = "\nYour previous attempt has duplicated points according to the required response format"
|
||||
println("\nERROR SQLLLM extractContent_dataframe() $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
|
||||
response = String(split(response, ", action_result")[1]) # in case LLM generate action_result key which it isn't supposed to
|
||||
response = strip(response)
|
||||
responsedict = nothing
|
||||
try
|
||||
_responsedict = JSON.parse(response)
|
||||
responsedict = GeneralUtils.dictify(_responsedict, keytype=String, sort_order=requiredKeys)
|
||||
catch
|
||||
println("\nERROR SQLLLM evaluator() failed to parse response: $response ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
|
||||
continue
|
||||
end
|
||||
|
||||
responsedict = GeneralUtils.textToDict(response, header; dictKey=dictkey, symbolkey=false)
|
||||
# check whether all answer's key points are in responsedict
|
||||
ispass, errormsg = GeneralUtils.checkAgentResponse_JSON(responsedict, requiredKeys)
|
||||
if !ispass
|
||||
errornote = errormsg
|
||||
println("\nERROR SQLLLM evaluator() $errornote --(not qualify response)> $responsedict ", @__FILE__, ":", @__LINE__, " $(Dates.now())\n")
|
||||
continue
|
||||
end
|
||||
|
||||
responsedict["selected_response_number"] = responsedict["selected_response_number"][1] # some time "6\nThe trajectories are incomplete" is generated but I only need the number.
|
||||
try
|
||||
responsedict["selected_response_number"] = parse(Int, responsedict["selected_response_number"]) # convert string "5" into integer 5
|
||||
try
|
||||
responsedict["selected_response_number"] = parse(Int, responsedict["selected_response_number"]) # convert string "5" into integer 5
|
||||
catch
|
||||
errornote = "In your previous attempt, Selected_response_number was not a number. It must be a number."
|
||||
println("\nERROR SQLLLM compareState() Attempt $attempt. $errornote ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
|
||||
continue
|
||||
end
|
||||
|
||||
println("\n~~~ compareState() ", @__FILE__, ":", @__LINE__, " $(Dates.now())")
|
||||
pprintln(Dict(responsedict))
|
||||
# println("\n~~~ compareState() ")
|
||||
# pprintln(Dict(responsedict))
|
||||
# println("---\n", @__FILE__, ":", @__LINE__)
|
||||
|
||||
return responsedict["selected_response_number"]
|
||||
end
|
||||
|
||||
+2
-2
@@ -137,7 +137,7 @@ end
|
||||
function insertSQLVectorDB(query::T1, SQL::T2; maxdistance::Integer=3) where {T1<:AbstractString, T2<:AbstractString}
|
||||
tablename = "sqlllm_decision_repository"
|
||||
# get embedding of the query
|
||||
# query = state[:thoughtHistory][:question]
|
||||
# query = state[:action_history][:question]
|
||||
df = findSimilarTextFromVectorDB(query, tablename,
|
||||
"function_input_embedding", executeSQLVectorDB)
|
||||
row, col = size(df)
|
||||
@@ -352,7 +352,7 @@ SELECT * FROM wine WHERE wine_type = 'red' AND country = 'France' AND sweetness
|
||||
# :evaluation =>
|
||||
# "The user's question is to search the database for wines that have a type of \"white\", are from \"France\", and have a sweetness level of 1. The thought is correct in identifying the conditions needed to filter the wine table. The action taken is to execute a SQL query to retrieve the desired data, which is also correct. The observation provides a search summary and two search results that match the user's question. Each result includes details about the wine such as ID, name, brand, manufacturer, region, country, type, grape variety, serving temperature, intensity, sweetness, tannin, and acidity.",
|
||||
# :accepted_as_answer => "Yes",
|
||||
# :thoughtHistory =>
|
||||
# :action_history =>
|
||||
# OrderedDict{String, Any}("question" => "Search the database for wine_type: white, country: France, sweetness: 1", "thought_1" => "The user wants to search the database for wines that have a type of \"white\", are from \"France\", and have a sweetness level of 1. To achieve this, we need to filter the wine table based on these conditions.", "action_name_1" => "GETDATA", "action_input_1" => "SELECT * FROM wine WHERE wine.wine_type = 'white' AND wine.country = 'France' AND wine.sweetness = 1;", "observation_1" => "\"Search summary: The resulting table represents wines.\\nSearch result: 1) wine_id: 5b6b6df9-d87c-4f33-8995-7249c2ecc917, wine_name: corton-charlemagne grand cru, brand: domaine des croix, manufacturer: domaine des croix, region: bourgogne, country: France, wine_type: white, grape_variety: cote de beaune blanc, serving_temperature: 11 to 13 Celsius, intensity: 4, sweetness: 1, tannin: missing, acidity: 3, fizziness: missing\\n2) wine_id: 1ad27d16-ef64-4907-acf1-40631630c143, wine_name: puligny-montrachet 1er cru 'les demoiselles', brand: amiot guy, manufacturer: amiot guy, region: bourgogne, country: France, wine_type: white, grape_variety: cote de beaune blanc, serving_temperature: 11 to 13 Celsius, intensity: 4, sweetness: 1, tannin: missing, acidity: 3, fizziness: missing\\n\\n\""),
|
||||
# :evaluationscore => 9,
|
||||
# :select => nothing,
|
||||
|
||||
Reference in New Issue
Block a user