fix: broadcast vectors for grad calculation #4179
Triggered via pull request
September 15, 2024 05:45
Status
Failure
Total duration
2m 49s
Artifacts
–
Annotations
34 errors and 16 warnings
Linux (nx, 1.15.4, 25.3)
test division rule computes gradient for constant denominator (Nx.Defn.GradTest)
|
Linux (nx, 1.15.4, 25.3)
test dot rule computes gradient for tensors on rhs (Nx.Defn.GradTest)
|
Linux (nx, 1.15.4, 25.3)
test take works with more dimensions (Nx.Defn.GradTest)
|
Linux (nx, 1.15.4, 25.3)
test triangular_solve computes the composed grad for tensor wrt b left_side: false (Nx.Defn.GradTest)
|
Linux (nx, 1.15.4, 25.3)
test window_min/max rule works with window max, inner function, same padding, stride (Nx.Defn.GradTest)
|
Linux (nx, 1.15.4, 25.3)
test put_slice computes gradient of operand (Nx.Defn.GradTest)
|
Linux (nx, 1.15.4, 25.3)
test dot rule computes gradient for tensors on lhs (Nx.Defn.GradTest)
|
Linux (nx, 1.15.4, 25.3)
test power rule computes gradient (Nx.Defn.GradTest)
|
Linux (nx, 1.15.4, 25.3)
test window sum rule computes the gradien with general padding (Nx.Defn.GradTest)
|
Linux (nx, 1.15.4, 25.3)
test product rule computes gradient for scalars (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test if computes gradient with sum (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test window sum rule works with nested window sums (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test erf computes the gradient (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test triangular_solve computes the grad for tensor wrt a lower: true, transform_a: :transpose (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test pad computes gradient (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test triangular_solve computes the composed grad for tensor wrt a lower: true (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test conv rule computes the gradient of both sides with arbitrary permutations (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test inverse trig family computes gradient of inverse trig functions (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test tokens computes token with grad (Nx.Defn.GradTest)
|
Linux (nx, 1.16.2, 25.3)
test atan2 rule computes gradient (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test conv rule computes the gradient for both sides, feature grouped (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test invert computes grad for complex tensor (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test broadcast computes gradient (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test log1p computes gradient (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test take works with more dimensions (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test sort computes gradient along axis (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test triangular_solve computes the composed grad for tensor wrt b left_side: false (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test as_type/bitcast as_type takes the real part when downcasting complex (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test indexed_add grad wrt to both source and target simultaneously (Nx.Defn.GradTest)
|
Windows (nx, 1.16.2, 25.2)
test concatenate computes grad for {1}-tensor (Nx.Defn.GradTest)
|
Linux (exla, 1.16.2, 25.3)
test if grad with tuple (EXLA.Defn.ExprTest)
|
Linux (exla, 1.16.2, 25.3)
Process completed with exit code 2.
|
Linux (exla, 1.15.4, 25.3)
test if grad with tuple (EXLA.Defn.ExprTest)
|
Linux (exla, 1.15.4, 25.3)
Process completed with exit code 2.
|
Linux (nx, 1.15.4, 25.3)
The following actions uses node12 which is deprecated and will be forced to run on node16: actions/checkout@v2. For more info: https://github.blog/changelog/2023-06-13-github-actions-all-actions-will-run-on-node16-instead-of-node12-by-default/
|
Linux (nx, 1.15.4, 25.3)
The following actions use a deprecated Node.js version and will be forced to run on node20: actions/checkout@v2, actions/cache@v3. For more info: https://github.blog/changelog/2024-03-07-github-actions-all-actions-will-run-on-node20-instead-of-node16-by-default/
|
Linux (nx, 1.16.2, 25.3)
The following actions uses node12 which is deprecated and will be forced to run on node16: actions/checkout@v2. For more info: https://github.blog/changelog/2023-06-13-github-actions-all-actions-will-run-on-node16-instead-of-node12-by-default/
|
Linux (nx, 1.16.2, 25.3)
The following actions use a deprecated Node.js version and will be forced to run on node20: actions/checkout@v2, actions/cache@v3. For more info: https://github.blog/changelog/2024-03-07-github-actions-all-actions-will-run-on-node20-instead-of-node16-by-default/
|
Linux (torchx, 1.16.2, 25.3)
The following actions uses node12 which is deprecated and will be forced to run on node16: actions/checkout@v2. For more info: https://github.blog/changelog/2023-06-13-github-actions-all-actions-will-run-on-node16-instead-of-node12-by-default/
|
Linux (torchx, 1.16.2, 25.3)
The following actions use a deprecated Node.js version and will be forced to run on node20: actions/checkout@v2, actions/cache@v3. For more info: https://github.blog/changelog/2024-03-07-github-actions-all-actions-will-run-on-node20-instead-of-node16-by-default/
|
Linux (torchx, 1.15.4, 25.3)
The following actions uses node12 which is deprecated and will be forced to run on node16: actions/checkout@v2. For more info: https://github.blog/changelog/2023-06-13-github-actions-all-actions-will-run-on-node16-instead-of-node12-by-default/
|
Linux (torchx, 1.15.4, 25.3)
The following actions use a deprecated Node.js version and will be forced to run on node20: actions/checkout@v2, actions/cache@v3. For more info: https://github.blog/changelog/2024-03-07-github-actions-all-actions-will-run-on-node20-instead-of-node16-by-default/
|
Windows (nx, 1.16.2, 25.2)
The following actions uses node12 which is deprecated and will be forced to run on node16: actions/checkout@v2. For more info: https://github.blog/changelog/2023-06-13-github-actions-all-actions-will-run-on-node16-instead-of-node12-by-default/
|
Windows (nx, 1.16.2, 25.2)
The following actions use a deprecated Node.js version and will be forced to run on node20: actions/checkout@v2, actions/cache@v3. For more info: https://github.blog/changelog/2024-03-07-github-actions-all-actions-will-run-on-node20-instead-of-node16-by-default/
|
Linux (exla, 1.16.2, 25.3)
The following actions uses node12 which is deprecated and will be forced to run on node16: actions/checkout@v2. For more info: https://github.blog/changelog/2023-06-13-github-actions-all-actions-will-run-on-node16-instead-of-node12-by-default/
|
Linux (exla, 1.16.2, 25.3)
The following actions use a deprecated Node.js version and will be forced to run on node20: actions/checkout@v2, actions/cache@v3. For more info: https://github.blog/changelog/2024-03-07-github-actions-all-actions-will-run-on-node20-instead-of-node16-by-default/
|
Linux (exla, 1.15.4, 25.3)
The following actions uses node12 which is deprecated and will be forced to run on node16: actions/checkout@v2. For more info: https://github.blog/changelog/2023-06-13-github-actions-all-actions-will-run-on-node16-instead-of-node12-by-default/
|
Linux (exla, 1.15.4, 25.3)
The following actions use a deprecated Node.js version and will be forced to run on node20: actions/checkout@v2, actions/cache@v3. For more info: https://github.blog/changelog/2024-03-07-github-actions-all-actions-will-run-on-node20-instead-of-node16-by-default/
|
Windows (torchx, 1.16.2, 25.2)
The following actions uses node12 which is deprecated and will be forced to run on node16: actions/checkout@v2. For more info: https://github.blog/changelog/2023-06-13-github-actions-all-actions-will-run-on-node16-instead-of-node12-by-default/
|
Windows (torchx, 1.16.2, 25.2)
The following actions use a deprecated Node.js version and will be forced to run on node20: actions/checkout@v2, actions/cache@v3. For more info: https://github.blog/changelog/2024-03-07-github-actions-all-actions-will-run-on-node20-instead-of-node16-by-default/
|