-
Notifications
You must be signed in to change notification settings - Fork 1.3k
skill: Add skills for problem solving (Clone of PR 428) #737
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Beep, boop 🤖, Hi, I'm @instructlab-bot and I'm going to help you with your pull request. Thanks for you contribution! 🎉 I support the following commands:
Note Results or Errors of these commands will be posted as a pull request check in the Checks section below Note Currently only maintainers belongs to [[taxonomy-triagers taxonomy-approvers taxonomy-maintainers labrador-org-maintainers instruct-lab-bot-maintainers]] teams are allowed to run these commands. |
1 similar comment
Beep, boop 🤖, Hi, I'm @instructlab-bot and I'm going to help you with your pull request. Thanks for you contribution! 🎉 I support the following commands:
Note Results or Errors of these commands will be posted as a pull request check in the Checks section below Note Currently only maintainers belongs to [[taxonomy-triagers taxonomy-approvers taxonomy-maintainers labrador-org-maintainers instruct-lab-bot-maintainers]] teams are allowed to run these commands. |
|
@instructlab-bot precheck |
Beep, boop 🤖, Generating test data for your PR with the job type: precheck. Your Job ID is 185. The results will be presented below in the pull request status box. This may take several minutes... |
Can you please fix the DCO on this PR? |
Beep, boop 🤖, Hi, I'm @instructlab-bot and I'm going to help you with your pull request. Thanks for you contribution! 🎉 I support the following commands:
Note Results or Errors of these commands will be posted as a pull request check in the Checks section below Note Currently only maintainers belongs to [[taxonomy-triagers taxonomy-approvers taxonomy-maintainers labrador-org-maintainers instruct-lab-bot-maintainers]] teams are allowed to run these commands. |
Beep, boop 🤖, Hi, I'm @instructlab-bot and I'm going to help you with your pull request. Thanks for you contribution! 🎉 I support the following commands:
Note Results or Errors of these commands will be posted as a pull request check in the Checks section below Note Currently only maintainers belongs to [[taxonomy-triagers taxonomy-approvers taxonomy-maintainers labrador-org-maintainers instruct-lab-bot-maintainers]] teams are allowed to run these commands. |
Done |
@instructlab-bot precheck |
Beep, boop 🤖, Generating test data for your PR with the job type: precheck. Your Job ID is 191. The results will be presented below in the pull request status box. This may take several minutes... |
Results for job ID: 191 using the model merlinite-7b! Results can be found here. |
Signed-off-by: Anil Kumar Vishnoi <avishnoi@redhat.com> Signed-off-by: Anil Vishnoi <vishnoianil@gmail.com> Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
Signed-off-by: Anil Kumar Vishnoi <avishnoi@redhat.com> Signed-off-by: Anil Vishnoi <vishnoianil@gmail.com> Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
Signed-off-by: Anil Kumar Vishnoi <avishnoi@redhat.com> Signed-off-by: Anil Vishnoi <vishnoianil@gmail.com> Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
Beep, boop 🤖, Hi, I'm @instructlab-bot and I'm going to help you with your pull request. Thanks for you contribution! 🎉 I support the following commands:
Note Results or Errors of these commands will be posted as a pull request check in the Checks section below Note Currently only maintainers belongs to [[taxonomy-triagers taxonomy-approvers taxonomy-maintainers labrador-org-maintainers instruct-lab-bot-maintainers]] teams are allowed to run these commands. |
Thank you for your submission. Here are your results: |
Thank you for your submission. Here are your results: |
) Reopening : **instructlab#428 **Describe the contribution to the taxonomy** - This PR adds skill to solve logical and inference riddles **Input given at the prompt & Response that was received** Attempted two riddles and also tried to question the model to fix the output Example 1: ``` >>> can you solve this riddle ? "Two turkeys in front of the turkey, two turkeys behind the turkey, tell me how many total turkeys?" ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ Certainly! This is an interesting riddle. The answer to "how many total turkeys" would be 10. Let me explain: │ │ │ │ The riddle states that there are two turkeys in front of the turkey and two turkeys behind the turkey. This implies that we have 1 │ │ turkey (the one in the middle) with 2 turkeys ahead and 2 more behind it. So, if we add the middle turkey to the ones in front and │ │ behind, we get a total of 10 turkeys. │ │ │ │ Here's a step-by-step breakdown: │ │ - There is 1 turkey in the middle (the one facing away from us in the initial description). │ │ - There are 2 turkeys in front of the middle turkey (the ones facing us in the initial description). │ │ - There are 2 turkeys behind the middle turkey (the ones facing away from us in the initial description). │ │ - Therefore, when we add the middle turkey to the ones in front and behind it, we get 1 (middle turkey) + 2 (front) + 2 (back) = 10 │ │ turkeys in total. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 9.359 seconds ─╯ >>> I think you got the basic math wrong, 1 +2 +2 equals to 5 and not 10. [S][default] ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ Ah, I see where the confusion might have arisen. Thank you for pointing it out! You are correct that 1 (middle turkey) + 2 (front) │ │ + 2 (back) should equal 5, not 10. The initial response included an error in the explanation. I apologize for any confusion caused. │ │ The correct answer to the riddle should be 5 turkeys in total, given the conditions provided. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 3.479 seconds ─╯ >>> But the answer is still not correct, it's not necessary that this riddle assumes one turkey is in the middle. [S][default] ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ You're right; the initial riddle description did not explicitly specify the positioning of the turkeys. Let's reconsider the riddle │ │ without assuming a middle turkey. Instead, let's assume there are three groups of turkeys: one group with 2 turkeys in front, │ │ another group with 2 turkeys behind, and a single turkey that is neither in front nor behind. In this case: │ │ - There are 2 turkeys in front of some unspecified turkey (let's call it A). │ │ - There are 2 turkeys behind some unspecified turkey (let's call it B). │ │ - There is 1 turkey (turkey C) that is not in front or behind any of the other turkeys. │ │ - Therefore, when we add all the turkeys together, we get 2 (front of A) + 2 (behind B) + 1 (C) = 5 turkeys in total. │ │ │ │ I apologize for any confusion caused earlier, and I appreciate your persistence in ensuring the correct answer is provided. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 9.162 seconds ─╯ ``` It got the basic math wrong in the first response, although after hinting the model about wrong math, it does corrected it, but It was not able to solve the riddle. Example 2: ``` >>> Tom's mother has 3 sons. The names of the two sons are dick and hardy, what's the name of the third son? [S][default] ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ Interesting question! However, based on the information provided, it is impossible to determine the name of Tom's third son, Dick. │ │ The names you provided, Dick and Hardy, do not match the typical naming patterns for sons, which often involve some form of │ │ repetition or variation of the father's name. Since there is no concrete information about the third son's name, I cannot provide a │ │ definitive answer. If there is additional information about Tom's family or naming traditions, I would be happy to revisit this │ │ question! │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 4.258 seconds ─╯ >>> Puzzle is not about finding sons of Tom, but sons of Tom's Mom. Can you take this information in consideration and try to solve the riddle? ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ Of course! Based on your new information, I will rephrase the question to reflect Tom's mother as the subject, and focus on finding │ │ her three sons: Dick, Hardy, and the third son, whose name is still unknown to us. Thank you for providing this clarification! │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 2.306 seconds ─╯ ``` Tried to assist the model to fix it's assumptions, but didn't solve the riddle. **Response that is now received instead** After training the model with the seed examples and answer, received the following responses from the models Example 1: ``` >>> can you solve the following riddle? Two turkeys in front [S][default] of the turkey, two turkeys behind the turkey, tell me how many total turkeys? ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ The answer is 5, because there are two turkeys in front, two turkeys behind, and one original turkey, making a total of five │ │ turkeys. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 1.254 seconds ─╯ >>> are you sure about it? you don't need to assume that you have turkey in the center, it can be at any position in the arrangement. ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ Yes, I am sure. The question asks for the total number of turkeys, which is five, regardless of their position or arrangement. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 1.034 seconds ─╯ >>> don't you think you can solve this riddle with 3 turkeys only? [S][default] ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ The question states that there are two turkeys in front and two turkeys behind, which adds up to four turkeys. However, there is │ │ one more turkey in the middle, making a total of five turkeys. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 1.677 seconds ─╯ ``` In the trained model, it got the math right but it was adamant to not learn from the given seed examples, and still giving wrong example. _I would appreciate any pointers here about how to make it better learn from the seed examples._ Example 2: ``` >>> can you solve the following riddle? Tom's mother has 3 sons. [S][default] The names of the two sons are dick and hardy, what's the name of the third son? ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ The answer is Tom, because the question asks for the name of the third son, and the names of the two other sons are Dick and Harry, │ │ making Tom the third son. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 1.360 seconds ─╯ >>> can you solve the following riddle? Ajay’s mother has 3 sons. [S][default] The names of the two sons are Vijay and Sanjay, what's the name of the third son? ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ The answer is Ajay, because the question asks for the name of the third son, and the names of the two other sons are Vijay and │ │ Sanjay, making Ajay the third son. ``` **Contribution checklist** - [x ] Contribution was tested with `lab generate` - [ x] No errors or warnings were produced by `lab generate` - [ x] All [commits are signed off](https://github.com/instruct-lab/taxonomy/blob/main/CONTRIBUTING.md#legal) (DCO) - [x ] The `qna.yaml` file was [linted](https://yamllint.com) and [prettified](https://onlineyamltools.com/prettify-yaml) ([yaml-validator](https://jsonformatter.org/yaml-validator) can do both) --------- Signed-off-by: Anil Kumar Vishnoi <avishnoi@redhat.com> Signed-off-by: Anil Vishnoi <vishnoianil@gmail.com> Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
) Reopening : **instructlab#428 **Describe the contribution to the taxonomy** - This PR adds skill to solve logical and inference riddles **Input given at the prompt & Response that was received** Attempted two riddles and also tried to question the model to fix the output Example 1: ``` >>> can you solve this riddle ? "Two turkeys in front of the turkey, two turkeys behind the turkey, tell me how many total turkeys?" ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ Certainly! This is an interesting riddle. The answer to "how many total turkeys" would be 10. Let me explain: │ │ │ │ The riddle states that there are two turkeys in front of the turkey and two turkeys behind the turkey. This implies that we have 1 │ │ turkey (the one in the middle) with 2 turkeys ahead and 2 more behind it. So, if we add the middle turkey to the ones in front and │ │ behind, we get a total of 10 turkeys. │ │ │ │ Here's a step-by-step breakdown: │ │ - There is 1 turkey in the middle (the one facing away from us in the initial description). │ │ - There are 2 turkeys in front of the middle turkey (the ones facing us in the initial description). │ │ - There are 2 turkeys behind the middle turkey (the ones facing away from us in the initial description). │ │ - Therefore, when we add the middle turkey to the ones in front and behind it, we get 1 (middle turkey) + 2 (front) + 2 (back) = 10 │ │ turkeys in total. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 9.359 seconds ─╯ >>> I think you got the basic math wrong, 1 +2 +2 equals to 5 and not 10. [S][default] ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ Ah, I see where the confusion might have arisen. Thank you for pointing it out! You are correct that 1 (middle turkey) + 2 (front) │ │ + 2 (back) should equal 5, not 10. The initial response included an error in the explanation. I apologize for any confusion caused. │ │ The correct answer to the riddle should be 5 turkeys in total, given the conditions provided. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 3.479 seconds ─╯ >>> But the answer is still not correct, it's not necessary that this riddle assumes one turkey is in the middle. [S][default] ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ You're right; the initial riddle description did not explicitly specify the positioning of the turkeys. Let's reconsider the riddle │ │ without assuming a middle turkey. Instead, let's assume there are three groups of turkeys: one group with 2 turkeys in front, │ │ another group with 2 turkeys behind, and a single turkey that is neither in front nor behind. In this case: │ │ - There are 2 turkeys in front of some unspecified turkey (let's call it A). │ │ - There are 2 turkeys behind some unspecified turkey (let's call it B). │ │ - There is 1 turkey (turkey C) that is not in front or behind any of the other turkeys. │ │ - Therefore, when we add all the turkeys together, we get 2 (front of A) + 2 (behind B) + 1 (C) = 5 turkeys in total. │ │ │ │ I apologize for any confusion caused earlier, and I appreciate your persistence in ensuring the correct answer is provided. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 9.162 seconds ─╯ ``` It got the basic math wrong in the first response, although after hinting the model about wrong math, it does corrected it, but It was not able to solve the riddle. Example 2: ``` >>> Tom's mother has 3 sons. The names of the two sons are dick and hardy, what's the name of the third son? [S][default] ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ Interesting question! However, based on the information provided, it is impossible to determine the name of Tom's third son, Dick. │ │ The names you provided, Dick and Hardy, do not match the typical naming patterns for sons, which often involve some form of │ │ repetition or variation of the father's name. Since there is no concrete information about the third son's name, I cannot provide a │ │ definitive answer. If there is additional information about Tom's family or naming traditions, I would be happy to revisit this │ │ question! │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 4.258 seconds ─╯ >>> Puzzle is not about finding sons of Tom, but sons of Tom's Mom. Can you take this information in consideration and try to solve the riddle? ╭──────────────────────────────────────────────────────── merlinite-7b-Q4_K_M ────────────────────────────────────────────────────────╮ │ Of course! Based on your new information, I will rephrase the question to reflect Tom's mother as the subject, and focus on finding │ │ her three sons: Dick, Hardy, and the third son, whose name is still unknown to us. Thank you for providing this clarification! │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 2.306 seconds ─╯ ``` Tried to assist the model to fix it's assumptions, but didn't solve the riddle. **Response that is now received instead** After training the model with the seed examples and answer, received the following responses from the models Example 1: ``` >>> can you solve the following riddle? Two turkeys in front [S][default] of the turkey, two turkeys behind the turkey, tell me how many total turkeys? ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ The answer is 5, because there are two turkeys in front, two turkeys behind, and one original turkey, making a total of five │ │ turkeys. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 1.254 seconds ─╯ >>> are you sure about it? you don't need to assume that you have turkey in the center, it can be at any position in the arrangement. ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ Yes, I am sure. The question asks for the total number of turkeys, which is five, regardless of their position or arrangement. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 1.034 seconds ─╯ >>> don't you think you can solve this riddle with 3 turkeys only? [S][default] ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ The question states that there are two turkeys in front and two turkeys behind, which adds up to four turkeys. However, there is │ │ one more turkey in the middle, making a total of five turkeys. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 1.677 seconds ─╯ ``` In the trained model, it got the math right but it was adamant to not learn from the given seed examples, and still giving wrong example. _I would appreciate any pointers here about how to make it better learn from the seed examples._ Example 2: ``` >>> can you solve the following riddle? Tom's mother has 3 sons. [S][default] The names of the two sons are dick and hardy, what's the name of the third son? ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ The answer is Tom, because the question asks for the name of the third son, and the names of the two other sons are Dick and Harry, │ │ making Tom the third son. │ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────── elapsed 1.360 seconds ─╯ >>> can you solve the following riddle? Ajay’s mother has 3 sons. [S][default] The names of the two sons are Vijay and Sanjay, what's the name of the third son? ╭───────────────────────────────────── ./ibm-merlinite-7b-mlx-q-fused-pt/ggml-model-Q4_K_M.gguf ──────────────────────────────────────╮ │ The answer is Ajay, because the question asks for the name of the third son, and the names of the two other sons are Vijay and │ │ Sanjay, making Ajay the third son. ``` **Contribution checklist** - [x ] Contribution was tested with `lab generate` - [ x] No errors or warnings were produced by `lab generate` - [ x] All [commits are signed off](https://github.com/instruct-lab/taxonomy/blob/main/CONTRIBUTING.md#legal) (DCO) - [x ] The `qna.yaml` file was [linted](https://yamllint.com) and [prettified](https://onlineyamltools.com/prettify-yaml) ([yaml-validator](https://jsonformatter.org/yaml-validator) can do both) --------- Signed-off-by: Anil Kumar Vishnoi <avishnoi@redhat.com> Signed-off-by: Anil Vishnoi <vishnoianil@gmail.com> Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
Reopening : #428
Describe the contribution to the taxonomy
Input given at the prompt & Response that was received
Attempted two riddles and also tried to question the model to fix the output
Example 1:
It got the basic math wrong in the first response, although after hinting the model about wrong math, it does corrected it, but It was not able to solve the riddle.
Example 2:
Tried to assist the model to fix it's assumptions, but didn't solve the riddle.
Response that is now received instead
After training the model with the seed examples and answer, received the following responses from the models
Example 1:
In the trained model, it got the math right but it was adamant to not learn from the given seed examples, and still giving wrong example. I would appreciate any pointers here about how to make it better learn from the seed examples.
Example 2:
Contribution checklist
lab generate
lab generate
qna.yaml
file was linted and prettified (yaml-validator can do both)