char-fork/llama.cpp
git clone https://git.t4t.associates/char-fork/llama.cpp
be20e7f49
char/aliceai-t5-moe
Debugging Tests Tips
How to run & execute or debug a specific test without anything else to keep the feedback loop short?
There is a script called debug-test.sh in the scripts folder whose parameter takes a REGEX and an optional test number.
For example, running the following command will output an interactive list from which you can select a test. It takes this form:
debug-test.sh [OPTION]... <test_regex> <test_number>
It will then build & run in the debugger for you.
To just execute a test and get back a PASS or FAIL message run:
./scripts/debug-test.sh test-tokenizer
To test in GDB use the -g flag to enable gdb test mode.
./scripts/debug-test.sh -g test-tokenizer# Once in the debugger, i.e. at the chevrons prompt, setting a breakpoint could be as follows: >>> bmain
To speed up the testing loop, if you know your test number you can just run it similar to below:
./scripts/debug-test.sh test 23
For further reference use debug-test.sh -h to print help.
How does the script work?
If you want to be able to use the concepts contained in the script separately, the important ones are briefly outlined below.
Step 1: Reset and Setup folder context
From base of this repository, let's create build-ci-debug as our build context.
rm -rf build-ci-debug&& mkdir build-ci-debug&& cd build-ci-debug
Step 2: Setup Build Environment and Compile Test Binaries
Setup and trigger a build under debug mode. You may adapt the arguments as needed, but in this case these are sane defaults.
cmake -DCMAKE_BUILD_TYPE=Debug -DLLAMA_CUDA=1 -DLLAMA_FATAL_WARNINGS=ON ..make -j
Step 3: Find all tests available that matches REGEX
The output of this command will give you the command & arguments needed to run GDB.
-R test-tokenizer: looks for all the test files namedtest-tokenizer*(R=Regex)-N: "show-only" disables test execution & shows test commands that you can feed to GDB.-V: Verbose Mode
ctest -R "test-tokenizer" -V -N
This may return output similar to below (focusing on key lines to pay attention to):
... 1: Test command: ~/llama.cpp/build-ci-debug/bin/test-tokenizer-0"~/llama.cpp/tests/../models/ggml-vocab-llama-spm.gguf" 1: Working Directory: .Labels: mainTest #1: test-tokenizer-0-llama-spm ... 4: Test command: ~/llama.cpp/build-ci-debug/bin/test-tokenizer-0"~/llama.cpp/tests/../models/ggml-vocab-falcon.gguf" 4: Working Directory: .Labels: mainTest #4: test-tokenizer-0-falcon ...
Step 4: Identify Test Command for Debugging
So for test #1 above we can tell these two pieces of relevant information:
- Test Binary:
~/llama.cpp/build-ci-debug/bin/test-tokenizer-0 - Test GGUF Model:
~/llama.cpp/tests/../models/ggml-vocab-llama-spm.gguf
Step 5: Run GDB on test command
Based on the ctest 'test command' report above we can then run a gdb session via this command below:
gdb --args ${Test Binary}${ Test GGUF Model}
Example:
gdb --args ~/llama.cpp/build-ci-debug/bin/test-tokenizer-0"~/llama.cpp/tests/../models/ggml-vocab-llama-spm.gguf"
1# Debugging Tests Tips 2 3## How to run & execute or debug a specific test without anything else to keep the feedback loop short? 4 5There is a script called debug-test.sh in the scripts folder whose parameter takes a REGEX and an optional test number. 6 7For example, running the following command will output an interactive list from which you can select a test. It takes this form: 8 9`debug-test.sh [OPTION]... <test_regex> <test_number>` 10 11It will then build & run in the debugger for you. 12 13To just execute a test and get back a PASS or FAIL message run: 14 15``` bash 16./scripts/debug-test.sh test-tokenizer 17``` 18 19To test in GDB use the `-g` flag to enable gdb test mode. 20 21``` bash 22./scripts/debug-test.sh -g test-tokenizer 23 24# Once in the debugger, i.e. at the chevrons prompt, setting a breakpoint could be as follows: 25>>> b main 26``` 27 28To speed up the testing loop, if you know your test number you can just run it similar to below: 29 30``` bash 31./scripts/debug-test.sh test 23 32``` 33 34For further reference use `debug-test.sh -h` to print help. 35 36 37 38### How does the script work? 39If you want to be able to use the concepts contained in the script separately, the important ones are briefly outlined below. 40 41#### Step 1: Reset and Setup folder context 42 43From base of this repository, let's create `build-ci-debug` as our build context. 44 45``` bash 46rm -rf build-ci-debug && mkdir build-ci-debug && cd build-ci-debug 47``` 48 49#### Step 2: Setup Build Environment and Compile Test Binaries 50 51Setup and trigger a build under debug mode. You may adapt the arguments as needed, but in this case these are sane defaults. 52 53``` bash 54cmake -DCMAKE_BUILD_TYPE=Debug -DLLAMA_CUDA=1 -DLLAMA_FATAL_WARNINGS=ON .. 55make -j 56``` 57 58#### Step 3: Find all tests available that matches REGEX 59 60The output of this command will give you the command & arguments needed to run GDB. 61 62* `-R test-tokenizer` : looks for all the test files named `test-tokenizer*` (R=Regex) 63* `-N` : "show-only" disables test execution & shows test commands that you can feed to GDB. 64* `-V` : Verbose Mode 65 66``` bash 67ctest -R "test-tokenizer" -V -N 68``` 69 70This may return output similar to below (focusing on key lines to pay attention to): 71 72``` bash 73... 741: Test command: ~/llama.cpp/build-ci-debug/bin/test-tokenizer-0 "~/llama.cpp/tests/../models/ggml-vocab-llama-spm.gguf" 751: Working Directory: . 76Labels: main 77Test #1: test-tokenizer-0-llama-spm 78... 794: Test command: ~/llama.cpp/build-ci-debug/bin/test-tokenizer-0 "~/llama.cpp/tests/../models/ggml-vocab-falcon.gguf" 804: Working Directory: . 81Labels: main 82Test #4: test-tokenizer-0-falcon 83... 84``` 85 86#### Step 4: Identify Test Command for Debugging 87 88So for test #1 above we can tell these two pieces of relevant information: 89* Test Binary: `~/llama.cpp/build-ci-debug/bin/test-tokenizer-0` 90* Test GGUF Model: `~/llama.cpp/tests/../models/ggml-vocab-llama-spm.gguf` 91 92#### Step 5: Run GDB on test command 93 94Based on the ctest 'test command' report above we can then run a gdb session via this command below: 95 96``` bash 97gdb --args ${ Test Binary} ${ Test GGUF Model} 98``` 99 100Example: 101 102``` bash 103gdb --args ~/llama.cpp/build-ci-debug/bin/test-tokenizer-0 "~/llama.cpp/tests/../models/ggml-vocab-llama-spm.gguf" 104```